metadata
license: apache-2.0
library_name: peft
tags:
- generated_from_trainer
datasets:
- common_voice_13_0
base_model: openai/whisper-tiny
model-index:
- name: arabic_speech_ASR_Whisper
results: []
arabic_speech_ASR_Whisper
This model is a fine-tuned version of openai/whisper-tiny on the common_voice_13_0 dataset. It achieves the following results on the evaluation set:
- Loss: 0.6463
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: 0.001
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.7591 | 1.0 | 300 | 0.7290 |
| 0.5736 | 2.0 | 600 | 0.6492 |
| 0.4507 | 3.0 | 900 | 0.6266 |
| 0.3456 | 4.0 | 1200 | 0.6130 |
| 0.2947 | 5.0 | 1500 | 0.6018 |
| 0.2153 | 6.0 | 1800 | 0.6061 |
| 0.1761 | 7.0 | 2100 | 0.6146 |
| 0.1185 | 8.0 | 2400 | 0.6290 |
| 0.0808 | 9.0 | 2700 | 0.6381 |
| 0.0643 | 10.0 | 3000 | 0.6463 |
Framework versions
- PEFT 0.9.0
- Transformers 4.38.2
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2