fake-news-bert
This model is a fine-tuned version of distilbert-base-uncased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.1353
- Accuracy: 0.9732
- Precision: 0.9934
- Recall: 0.9527
- F1: 0.9726
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: 16
- eval_batch_size: 32
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 4
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|---|---|---|---|---|---|---|---|
| 0.1432 | 1.0 | 317 | 0.1278 | 0.9637 | 0.9803 | 0.9462 | 0.9630 |
| 0.0668 | 2.0 | 634 | 0.1583 | 0.9573 | 0.9833 | 0.9304 | 0.9561 |
| 0.0150 | 3.0 | 951 | 0.1253 | 0.9763 | 0.9839 | 0.9684 | 0.9761 |
| 0.0221 | 4.0 | 1268 | 0.1278 | 0.9716 | 0.9656 | 0.9778 | 0.9717 |
Framework versions
- Transformers 5.16.1
- Pytorch 2.11.0+cu128
- Datasets 4.8.5
- Tokenizers 0.23.1