CS221-bert-base-uncased-finetuned-semeval-NT-rus
This model is a fine-tuned version of google-bert/bert-base-uncased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.1875
- F1: 0.8610
- Roc Auc: 0.9156
- Accuracy: 0.8228
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
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 | Roc Auc | Accuracy |
|---|---|---|---|---|---|---|
| 0.2467 | 1.0 | 268 | 0.1972 | 0.8117 | 0.8645 | 0.7780 |
| 0.132 | 2.0 | 536 | 0.1482 | 0.8393 | 0.8852 | 0.8078 |
| 0.098 | 3.0 | 804 | 0.1554 | 0.8405 | 0.9010 | 0.8060 |
| 0.0563 | 4.0 | 1072 | 0.1658 | 0.8515 | 0.9019 | 0.8209 |
| 0.0247 | 5.0 | 1340 | 0.1699 | 0.8475 | 0.9075 | 0.8097 |
| 0.0187 | 6.0 | 1608 | 0.1709 | 0.8545 | 0.9050 | 0.8190 |
| 0.016 | 7.0 | 1876 | 0.1771 | 0.8457 | 0.9030 | 0.7985 |
| 0.0079 | 8.0 | 2144 | 0.1851 | 0.8568 | 0.9089 | 0.8228 |
| 0.0145 | 9.0 | 2412 | 0.1875 | 0.8610 | 0.9156 | 0.8228 |
| 0.0041 | 10.0 | 2680 | 0.1942 | 0.8542 | 0.9041 | 0.8265 |
| 0.0033 | 11.0 | 2948 | 0.1964 | 0.8581 | 0.9125 | 0.8265 |
| 0.003 | 12.0 | 3216 | 0.1979 | 0.8565 | 0.9105 | 0.8228 |
Framework versions
- Transformers 4.47.1
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0