Configuration Parsing
Warning:
In adapter_config.json: "peft.task_type" must be a string
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: 0.6028
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 |
|---|---|---|---|
| 0.7473 | 1.0 | 300 | 0.6873 |
| 0.5487 | 2.0 | 600 | 0.6098 |
| 0.4391 | 3.0 | 900 | 0.5909 |
| 0.3461 | 4.0 | 1200 | 0.5766 |
| 0.3315 | 5.0 | 1500 | 0.5552 |
| 0.2339 | 6.0 | 1800 | 0.5694 |
| 0.1609 | 7.0 | 2100 | 0.5699 |
| 0.1236 | 8.0 | 2400 | 0.5879 |
| 0.0898 | 9.0 | 2700 | 0.5915 |
| 0.0607 | 10.0 | 3000 | 0.6028 |
Framework versions
- PEFT 0.9.0
- Transformers 4.38.2
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2
- Downloads last month
- 2
Model tree for Kaizukony/arabic_speech_ASR_Whisper
Base model
openai/whisper-tiny