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: 0.0037
- Wer: 0.0571
- Cer: 0.0223
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: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- 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: 25
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| 0.0078 | 1.0 | 282 | 0.0037 | 0.0529 | 0.0209 |
| 0.0047 | 2.0 | 564 | 0.0040 | 0.0550 | 0.0208 |
| 0.0018 | 3.0 | 846 | 0.0050 | 0.0646 | 0.0261 |
| 0.0013 | 4.0 | 1128 | 0.0053 | 0.0594 | 0.0224 |
| 0.0009 | 5.0 | 1410 | 0.0062 | 0.0659 | 0.0241 |
| 0.0008 | 6.0 | 1692 | 0.0066 | 0.0659 | 0.0257 |
| 0.0008 | 7.0 | 1974 | 0.0068 | 0.0626 | 0.0243 |
| 0.0006 | 8.0 | 2256 | 0.0072 | 0.0615 | 0.0223 |
| 0.0006 | 9.0 | 2538 | 0.0075 | 0.0668 | 0.0256 |
| 0.0003 | 10.0 | 2820 | 0.0077 | 0.0643 | 0.0238 |
| 0.0003 | 11.0 | 3102 | 0.0082 | 0.0577 | 0.0211 |
| 0.0002 | 12.0 | 3384 | 0.0090 | 0.0643 | 0.0237 |
| 0.0002 | 13.0 | 3666 | 0.0100 | 0.0637 | 0.0222 |
| 0.0001 | 14.0 | 3948 | 0.0100 | 0.0615 | 0.0229 |
| 0.0001 | 15.0 | 4230 | 0.0104 | 0.0603 | 0.0232 |
| 0.0001 | 16.0 | 4512 | 0.0111 | 0.0606 | 0.0216 |
| 0.0001 | 17.0 | 4794 | 0.0115 | 0.0614 | 0.0211 |
| 0.0 | 18.0 | 5076 | 0.0115 | 0.0570 | 0.0196 |
| 0.0 | 19.0 | 5358 | 0.0121 | 0.0614 | 0.0225 |
| 0.0 | 20.0 | 5640 | 0.0119 | 0.0595 | 0.0219 |
| 0.0 | 21.0 | 5922 | 0.0121 | 0.0588 | 0.0209 |
| 0.0 | 22.0 | 6204 | 0.0121 | 0.0597 | 0.0215 |
| 0.0 | 23.0 | 6486 | 0.0121 | 0.0601 | 0.0215 |
| 0.0 | 24.0 | 6768 | 0.0126 | 0.0609 | 0.0234 |
| 0.0 | 24.9130 | 7025 | 0.0121 | 0.0588 | 0.0204 |
Framework versions
- Transformers 4.51.1
- Pytorch 2.5.1+cu124
- Datasets 3.5.0
- Tokenizers 0.21.0