metadata
language:
- ar
license: apache-2.0
base_model: tarteel-ai/whisper-base-ar-quran
tags:
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_11_0
metrics:
- wer
model-index:
- name: raghaddd
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Zolfa Dataset
type: mozilla-foundation/common_voice_11_0
args: 'config: ar, split: test'
metrics:
- name: Wer
type: wer
value: 10.344827586206897
raghaddd
This model is a fine-tuned version of tarteel-ai/whisper-base-ar-quran on the Zolfa Dataset dataset. It achieves the following results on the evaluation set:
- Loss: 0.0160
- Wer: 10.3448
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: 2
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.382 | 1.0 | 21 | 0.0234 | 6.8966 |
| 0.0151 | 2.0 | 42 | 0.0056 | 0.0 |
| 0.0033 | 3.0 | 63 | 0.0066 | 0.0 |
| 0.0014 | 4.0 | 84 | 0.0112 | 10.3448 |
| 0.001 | 5.0 | 105 | 0.0113 | 3.4483 |
| 0.0005 | 6.0 | 126 | 0.0141 | 3.4483 |
| 0.0004 | 7.0 | 147 | 0.0140 | 10.3448 |
| 0.0005 | 8.0 | 168 | 0.0165 | 10.3448 |
| 0.0003 | 9.0 | 189 | 0.0145 | 10.3448 |
| 0.0003 | 10.0 | 210 | 0.0160 | 10.3448 |
Framework versions
- Transformers 4.40.2
- Pytorch 2.2.1+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1