metadata
language:
- ar
license: apache-2.0
base_model: openai/whisper-small
tags:
- generated_from_trainer
datasets:
- ayoubkirouane/darija-stt-mix
metrics:
- wer
model-index:
- name: Whisper Small Egyptian Arabic
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Darija STT Mix
type: ayoubkirouane/darija-stt-mix
metrics:
- name: Wer
type: wer
value: 40.95437860513896
Whisper Small Egyptian Arabic
This model is a fine-tuned version of openai/whisper-small on the Darija STT Mix dataset. It achieves the following results on the evaluation set:
- Loss: 0.6220
- Wer Ortho: 45.3540
- Wer: 40.9544
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: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 200
- training_steps: 1500
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer Ortho | Wer |
|---|---|---|---|---|---|
| 0.7241 | 0.7587 | 500 | 0.7187 | 55.7053 | 51.3005 |
| 0.411 | 1.5175 | 1000 | 0.6399 | 49.0351 | 44.4887 |
| 0.2807 | 2.2762 | 1500 | 0.6220 | 45.3540 | 40.9544 |
Framework versions
- Transformers 4.42.4
- Pytorch 2.4.0+cu121
- Datasets 2.21.0
- Tokenizers 0.19.1