metadata
library_name: transformers
language:
- ar
license: apache-2.0
base_model: openai/whisper-base
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: Whisper base AR - BH
results: []
Whisper base AR - BH
This model is a fine-tuned version of openai/whisper-base on the quran-ayat-speech-to-text-segments dataset. It achieves the following results on the evaluation set:
- Loss: 0.0231
- Wer: 0.2331
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: 8
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- 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 |
|---|---|---|---|---|
| 0.017 | 1.3333 | 100 | 0.0233 | 0.2610 |
| 0.0148 | 2.6667 | 200 | 0.0216 | 0.2495 |
| 0.0094 | 4.0 | 300 | 0.0204 | 0.2438 |
| 0.0048 | 5.3333 | 400 | 0.0203 | 0.2409 |
| 0.0027 | 6.6667 | 500 | 0.0206 | 0.2421 |
| 0.0017 | 8.0 | 600 | 0.0213 | 0.2491 |
| 0.0007 | 9.3333 | 700 | 0.0218 | 0.2384 |
| 0.0005 | 10.6667 | 800 | 0.0221 | 0.2368 |
| 0.0003 | 12.0 | 900 | 0.0224 | 0.2372 |
| 0.0003 | 13.3333 | 1000 | 0.0226 | 0.2347 |
| 0.0003 | 14.6667 | 1100 | 0.0228 | 0.2327 |
| 0.0002 | 16.0 | 1200 | 0.0229 | 0.2335 |
| 0.0002 | 17.3333 | 1300 | 0.0230 | 0.2331 |
| 0.0002 | 18.6667 | 1400 | 0.0231 | 0.2331 |
| 0.0002 | 20.0 | 1500 | 0.0231 | 0.2335 |
Framework versions
- Transformers 4.46.3
- Pytorch 2.4.0
- Datasets 3.1.0
- Tokenizers 0.20.3