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: 0.0253
- Wer: 0.1830
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
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- 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 |
|---|---|---|---|---|
| 0.0158 | 1.0 | 100 | 0.0179 | 0.2185 |
| 0.0131 | 2.0 | 200 | 0.0161 | 0.2154 |
| 0.0105 | 3.0 | 300 | 0.0148 | 0.2059 |
| 0.0066 | 4.0 | 400 | 0.0144 | 0.2011 |
| 0.0046 | 5.0 | 500 | 0.0145 | 0.2011 |
| 0.0021 | 6.0 | 600 | 0.0150 | 0.1985 |
| 0.0012 | 7.0 | 700 | 0.0154 | 0.1945 |
| 0.0004 | 8.0 | 800 | 0.0161 | 0.1890 |
| 0.0002 | 9.0 | 900 | 0.0169 | 0.1894 |
| 0.0001 | 10.0 | 1000 | 0.0177 | 0.1899 |
| 0.0 | 11.0 | 1100 | 0.0185 | 0.1842 |
| 0.0 | 12.0 | 1200 | 0.0193 | 0.1850 |
| 0.0 | 13.0 | 1300 | 0.0203 | 0.1839 |
| 0.0 | 14.0 | 1400 | 0.0214 | 0.1844 |
| 0.0 | 15.0 | 1500 | 0.0224 | 0.1804 |
| 0.0 | 16.0 | 1600 | 0.0234 | 0.1833 |
| 0.0 | 17.0 | 1700 | 0.0243 | 0.1818 |
| 0.0 | 18.0 | 1800 | 0.0253 | 0.1830 |
| 0.0 | 19.0 | 1900 | 0.0259 | 0.1854 |
| 0.0 | 20.0 | 2000 | 0.0261 | 0.1854 |
Framework versions
- Transformers 4.46.3
- Pytorch 2.4.0
- Datasets 3.1.0
- Tokenizers 0.20.3