Paper-Matcher-bert
This model is a fine-tuned version of bert-base-uncased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.5883
- Accuracy: 0.87
- Macro F1: 0.8699
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
- lr_scheduler_warmup_steps: 75
- num_epochs: 6
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Macro F1 |
|---|---|---|---|---|---|
| 0.3503 | 1.0 | 125 | 0.4379 | 0.83 | 0.8294 |
| 0.2971 | 2.0 | 250 | 0.4430 | 0.855 | 0.8549 |
| 0.2569 | 3.0 | 375 | 0.5001 | 0.87 | 0.87 |
| 0.1983 | 4.0 | 500 | 0.5181 | 0.855 | 0.8547 |
| 0.1342 | 5.0 | 625 | 0.5464 | 0.86 | 0.8598 |
| 0.1238 | 6.0 | 750 | 0.5883 | 0.87 | 0.8699 |
Framework versions
- Transformers 5.18.0
- Pytorch 2.11.0+cu130
- Datasets 4.8.5
- Tokenizers 0.23.2