<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
whisper-small-ko-E
This model is a fine-tuned version of openai/whisper-small on the aihub elder over 70 dataset. It achieves the following results on the evaluation set:
- Loss: 0.2003
- Cer: 4.6223
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: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 0.2246 | 0.3 | 100 | 0.2507 | 6.3431 |
| 0.2129 | 0.59 | 200 | 0.2300 | 6.3668 |
| 0.1896 | 0.89 | 300 | 0.2029 | 5.5302 |
| 0.0692 | 1.19 | 400 | 0.2010 | 5.4768 |
| 0.0636 | 1.48 | 500 | 0.2063 | 5.3700 |
| 0.0625 | 1.78 | 600 | 0.1999 | 5.2157 |
| 0.0239 | 2.07 | 700 | 0.1919 | 4.7291 |
| 0.026 | 2.37 | 800 | 0.2023 | 4.8181 |
| 0.0261 | 2.67 | 900 | 0.1950 | 4.6935 |
| 0.0241 | 2.96 | 1000 | 0.1900 | 4.8181 |
| 0.0085 | 3.26 | 1100 | 0.1997 | 4.6935 |
| 0.0072 | 3.56 | 1200 | 0.1951 | 4.5333 |
| 0.0087 | 3.85 | 1300 | 0.1990 | 4.7766 |
| 0.0031 | 4.15 | 1400 | 0.2006 | 4.6461 |
| 0.0026 | 4.44 | 1500 | 0.2004 | 4.7054 |
| 0.0024 | 4.74 | 1600 | 0.2003 | 4.6223 |
Framework versions
- Transformers 4.34.0.dev0
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.13.3
