whisper-small-ko-E50_Y_freq_speed-SA
This model is a fine-tuned version of openai/whisper-small on the aihub Y dialogue dataset. It achieves the following results on the evaluation set:
- Loss: 0.1737
- Cer: 5.7155
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: 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: 2
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Cer |
---|---|---|---|---|
0.4988 | 0.13 | 100 | 0.2885 | 7.1840 |
0.3371 | 0.26 | 200 | 0.2180 | 5.7977 |
0.2889 | 0.39 | 300 | 0.2138 | 6.25 |
0.258 | 0.52 | 400 | 0.2019 | 5.7977 |
0.2357 | 0.64 | 500 | 0.1965 | 5.4688 |
0.219 | 0.77 | 600 | 0.1865 | 6.2852 |
0.2119 | 0.9 | 700 | 0.1832 | 5.3160 |
0.1416 | 1.03 | 800 | 0.1778 | 5.1692 |
0.126 | 1.16 | 900 | 0.1813 | 5.0576 |
0.1346 | 1.29 | 1000 | 0.1778 | 5.0047 |
0.1205 | 1.42 | 1100 | 0.1778 | 4.9518 |
0.1121 | 1.55 | 1200 | 0.1745 | 4.9283 |
0.1259 | 1.68 | 1300 | 0.1736 | 6.1149 |
0.1128 | 1.81 | 1400 | 0.1739 | 5.6978 |
0.1027 | 1.93 | 1500 | 0.1737 | 5.7155 |
Framework versions
- Transformers 4.37.0.dev0
- Pytorch 2.1.0+cu121
- Datasets 2.16.0
- Tokenizers 0.15.0
- Downloads last month
- 5