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 - BA
results: []
Whisper base AR - BA
This model is a fine-tuned version of openai/whisper-base on the quran-ayat-speech-to-text dataset. It achieves the following results on the evaluation set:
- Loss: 0.1257
- Wer: 0.3013
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Use OptimizerNames.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: 15
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 6.5463 | 1.0 | 469 | 0.1230 | 0.3954 |
| 5.2743 | 2.0 | 938 | 0.1311 | 0.3571 |
| 3.5152 | 3.0 | 1407 | 0.1228 | 0.3684 |
| 2.7744 | 4.0 | 1876 | 0.1212 | 0.3571 |
| 2.2344 | 5.0 | 2345 | 0.1207 | 0.3227 |
| 1.8514 | 6.0 | 2814 | 0.1217 | 0.3179 |
| 1.6307 | 7.0 | 3283 | 0.1182 | 0.3561 |
| 1.35 | 8.0 | 3752 | 0.1187 | 0.2904 |
| 1.3541 | 9.0 | 4221 | 0.1135 | 0.2999 |
| 1.2367 | 10.0 | 4690 | 0.1148 | 0.2886 |
| 1.1444 | 11.0 | 5159 | 0.1147 | 0.3009 |
| 0.9895 | 12.0 | 5628 | 0.1139 | 0.2882 |
| 0.9394 | 13.0 | 6097 | 0.1126 | 0.3167 |
| 0.9161 | 14.0 | 6566 | 0.1125 | 0.2967 |
| 0.8473 | 14.9685 | 7020 | 0.1130 | 0.2929 |
Framework versions
- Transformers 4.51.3
- Pytorch 2.6.0+cu124
- Datasets 3.6.0
- Tokenizers 0.21.1