xlm-roberta-baseFullTune
This model is a fine-tuned version of xlm-roberta-base on SIRIS-Lab/citation-parser-ENTITY dataset. It achieves the following results on the evaluation set:
- Loss: 0.0905
- Precision: 0.9605
- Recall: 0.9675
- F1: 0.9640
- Accuracy: 0.9852
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: 32
- 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: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 1.6802 | 1.0 | 68 | 0.3909 | 0.6138 | 0.6296 | 0.6216 | 0.9132 |
| 0.3432 | 2.0 | 136 | 0.1687 | 0.8981 | 0.9357 | 0.9165 | 0.9670 |
| 0.0845 | 3.0 | 204 | 0.1430 | 0.9298 | 0.9537 | 0.9416 | 0.9742 |
| 0.0689 | 4.0 | 272 | 0.1330 | 0.9403 | 0.9627 | 0.9514 | 0.9773 |
| 0.0497 | 5.0 | 340 | 0.1212 | 0.9548 | 0.9653 | 0.9600 | 0.9782 |
| 0.0400 | 6.0 | 408 | 0.1253 | 0.9506 | 0.9653 | 0.9579 | 0.9784 |
| 0.0363 | 7.0 | 476 | 0.1189 | 0.9543 | 0.9659 | 0.9601 | 0.9791 |
| 0.0333 | 8.0 | 544 | 0.1219 | 0.9592 | 0.9666 | 0.9628 | 0.9783 |
| 0.0257 | 9.0 | 612 | 0.1244 | 0.9573 | 0.9659 | 0.9616 | 0.9781 |
| 0.0299 | 10.0 | 680 | 0.1247 | 0.9561 | 0.9659 | 0.9610 | 0.9776 |
Framework versions
- Transformers 5.13.1
- Pytorch 2.11.0+cu128
- Datasets 4.0.0
- Tokenizers 0.22.2