<!-- 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. -->
Yepes_2e-05_31_03
This model is a fine-tuned version of microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0955
- Precision: 0.6940
- Recall: 0.5159
- F1: 0.5918
- Accuracy: 0.9824
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
- training_steps: 500
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
0.7814 | 1.92 | 25 | 0.2044 | 0.0 | 0.0 | 0.0 | 0.9704 |
0.1936 | 3.85 | 50 | 0.1853 | 0.0 | 0.0 | 0.0 | 0.9704 |
0.1651 | 5.77 | 75 | 0.1260 | 0.1667 | 0.0714 | 0.1 | 0.9714 |
0.1129 | 7.69 | 100 | 0.1127 | 0.3532 | 0.2354 | 0.2825 | 0.9761 |
0.0906 | 9.62 | 125 | 0.1029 | 0.4878 | 0.3175 | 0.3846 | 0.9788 |
0.0773 | 11.54 | 150 | 0.0990 | 0.5059 | 0.3386 | 0.4057 | 0.9797 |
0.0621 | 13.46 | 175 | 0.0908 | 0.5486 | 0.4180 | 0.4745 | 0.9793 |
0.0568 | 15.38 | 200 | 0.0875 | 0.6212 | 0.4339 | 0.5109 | 0.9811 |
0.0462 | 17.31 | 225 | 0.0900 | 0.6413 | 0.4683 | 0.5413 | 0.9817 |
0.0403 | 19.23 | 250 | 0.0936 | 0.6797 | 0.4603 | 0.5489 | 0.9823 |
0.0372 | 21.15 | 275 | 0.0895 | 0.6516 | 0.4947 | 0.5624 | 0.9819 |
0.0333 | 23.08 | 300 | 0.0925 | 0.6643 | 0.4974 | 0.5688 | 0.9823 |
0.0285 | 25.0 | 325 | 0.0962 | 0.6871 | 0.5053 | 0.5823 | 0.9822 |
0.0262 | 26.92 | 350 | 0.0955 | 0.6940 | 0.5159 | 0.5918 | 0.9824 |
Framework versions
- Transformers 4.27.4
- Pytorch 1.13.1+cu116
- Datasets 2.11.0
- Tokenizers 0.13.2