<!-- 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. -->
korean_sentiment_analysis_dataset3_best
This model is a fine-tuned version of klue/roberta-large on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.6989
- Micro f1 score: 76.6383
- Auprc: 81.5157
- Accuracy: 0.7664
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: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 3.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Micro f1 score | Auprc | Accuracy |
|---|---|---|---|---|---|---|
| 0.7997 | 1.0 | 5080 | 0.6822 | 74.7769 | 79.4361 | 0.7478 |
| 0.4544 | 2.0 | 10160 | 0.6608 | 76.7429 | 81.1265 | 0.7674 |
| 0.5702 | 3.0 | 15240 | 0.6989 | 76.6383 | 81.5157 | 0.7664 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.6.0
- Datasets 2.7.1
- Tokenizers 0.13.2
