1008model
This model is a fine-tuned version of beomi/kcbert-base on the unsmile_data dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1289
- Lrap: 0.8808
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: 64
- eval_batch_size: 64
- 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 |
Lrap |
| No log |
1.0 |
235 |
0.1471 |
0.8580 |
| No log |
2.0 |
470 |
0.1265 |
0.8766 |
| 0.1688 |
3.0 |
705 |
0.1221 |
0.8823 |
| 0.1688 |
4.0 |
940 |
0.1272 |
0.8825 |
| 0.0772 |
5.0 |
1175 |
0.1289 |
0.8808 |
Framework versions
- Transformers 4.44.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.1
- Tokenizers 0.19.1
