<!-- 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. -->
PubMedBERT-Large-LitCovid-v1.3.1
This model is a fine-tuned version of microsoft/BiomedNLP-PubMedBERT-large-uncased-abstract on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.2334
- Hamming loss: 0.0114
- F1 micro: 0.8967
- F1 macro: 0.5317
- F1 weighted: 0.9010
- F1 samples: 0.9021
- Precision micro: 0.8733
- Precision macro: 0.4705
- Precision weighted: 0.8829
- Precision samples: 0.9034
- Recall micro: 0.9214
- Recall macro: 0.6606
- Recall weighted: 0.9214
- Recall samples: 0.9340
- Roc Auc: 0.9569
- Accuracy: 0.7445
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: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
Training results
Training Loss | Epoch | Step | Validation Loss | Hamming loss | F1 micro | F1 macro | F1 weighted | F1 samples | Precision micro | Precision macro | Precision weighted | Precision samples | Recall micro | Recall macro | Recall weighted | Recall samples | Roc Auc | Accuracy |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2.4299 | 1.0 | 9086 | 0.8402 | 0.0183 | 0.8440 | 0.3442 | 0.8744 | 0.8786 | 0.7770 | 0.2832 | 0.8384 | 0.8690 | 0.9236 | 0.6237 | 0.9236 | 0.9363 | 0.9543 | 0.6930 |
1.388 | 2.0 | 18172 | 0.7233 | 0.0161 | 0.8613 | 0.3836 | 0.8886 | 0.8913 | 0.8016 | 0.3164 | 0.8565 | 0.8827 | 0.9306 | 0.6955 | 0.9306 | 0.9426 | 0.9588 | 0.7163 |
0.9723 | 3.0 | 27258 | 0.8506 | 0.0132 | 0.8829 | 0.4584 | 0.8970 | 0.8963 | 0.8434 | 0.3870 | 0.8735 | 0.8924 | 0.9264 | 0.6941 | 0.9264 | 0.9380 | 0.9583 | 0.7313 |
0.7713 | 4.0 | 36344 | 1.0207 | 0.0119 | 0.8923 | 0.5060 | 0.8994 | 0.8990 | 0.8635 | 0.4387 | 0.8795 | 0.8976 | 0.9230 | 0.6758 | 0.9230 | 0.9349 | 0.9574 | 0.7389 |
0.3859 | 5.0 | 45430 | 1.2334 | 0.0114 | 0.8967 | 0.5317 | 0.9010 | 0.9021 | 0.8733 | 0.4705 | 0.8829 | 0.9034 | 0.9214 | 0.6606 | 0.9214 | 0.9340 | 0.9569 | 0.7445 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.0
- Datasets 2.1.0
- Tokenizers 0.13.3