Whisper tiny mozilla-foundation/common_voice_11_0 - Huang Jordan
This model is a fine-tuned version of openai/whisper-tiny on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:
- Loss: 0.4430
- Cer: 22.6389
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
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 200
- training_steps: 2000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 0.4622 | 0.7092 | 500 | 0.4827 | 24.3222 |
| 0.3287 | 1.4184 | 1000 | 0.4569 | 22.5015 |
| 0.2613 | 2.1277 | 1500 | 0.4454 | 22.3270 |
| 0.24 | 2.8369 | 2000 | 0.4430 | 22.6389 |
Framework versions
- Transformers 4.40.0
- Pytorch 2.2.2+cu121
- Datasets 2.19.0
- Tokenizers 0.19.1
- Downloads last month
- 8