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: 0.4170
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: 6
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.6558 | 1.0 | 1334 | 0.6054 |
| 0.4922 | 2.0 | 2668 | 0.5290 |
| 0.4324 | 3.0 | 4002 | 0.4767 |
| 0.3572 | 4.0 | 5336 | 0.4277 |
| 0.2179 | 5.0 | 6670 | 0.4170 |
Framework versions
- PEFT 0.9.0
- Transformers 4.38.1
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2