<!-- 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. -->
distilroberta-base-mapa_coarse-ner
This model is a fine-tuned version of distilroberta-base on the lextreme dataset. It achieves the following results on the evaluation set:
- Loss: 0.1020
- Precision: 0.7441
- Recall: 0.5805
- F1: 0.6522
- Accuracy: 0.9872
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: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 15
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.0343 | 1.0 | 1739 | 0.0694 | 0.6342 | 0.5205 | 0.5718 | 0.9841 |
| 0.0263 | 2.0 | 3478 | 0.0705 | 0.7961 | 0.5235 | 0.6317 | 0.9865 |
| 0.0183 | 3.0 | 5217 | 0.0670 | 0.7417 | 0.5313 | 0.6191 | 0.9864 |
| 0.015 | 4.0 | 6956 | 0.0632 | 0.7237 | 0.5850 | 0.6470 | 0.9869 |
| 0.0137 | 5.0 | 8695 | 0.0663 | 0.7311 | 0.6064 | 0.6629 | 0.9872 |
| 0.011 | 6.0 | 10434 | 0.0703 | 0.7163 | 0.5877 | 0.6457 | 0.9868 |
| 0.0096 | 7.0 | 12173 | 0.0799 | 0.7511 | 0.5676 | 0.6466 | 0.9871 |
| 0.0071 | 8.0 | 13912 | 0.0770 | 0.7386 | 0.5640 | 0.6396 | 0.9868 |
| 0.0068 | 9.0 | 15651 | 0.0827 | 0.7285 | 0.5674 | 0.6379 | 0.9868 |
| 0.0057 | 10.0 | 17390 | 0.0897 | 0.7611 | 0.5719 | 0.6531 | 0.9872 |
| 0.0053 | 11.0 | 19129 | 0.0940 | 0.7614 | 0.5627 | 0.6471 | 0.9871 |
| 0.004 | 12.0 | 20868 | 0.0874 | 0.7184 | 0.6084 | 0.6588 | 0.9873 |
| 0.0035 | 13.0 | 22607 | 0.0986 | 0.7513 | 0.5766 | 0.6525 | 0.9872 |
| 0.003 | 14.0 | 24346 | 0.1012 | 0.7396 | 0.5805 | 0.6505 | 0.9871 |
| 0.0026 | 15.0 | 26085 | 0.1020 | 0.7441 | 0.5805 | 0.6522 | 0.9872 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.13.1+cu117
- Datasets 2.9.0
- Tokenizers 0.13.2
