<!-- 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. -->
BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext-finetuned-ner
This model is a fine-tuned version of microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext fine tuned on the CORD-19 dataset. It achieves the following results on the evaluation set:
- Accuracy: 0.8988
- Loss: 0.4018
- Precision: 0.63
- Recall: 0.4701
- F1: 0.5385
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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 4 | 0.4300 | 0.6667 | 0.3134 | 0.4264 | 0.8774 |
| No log | 2.0 | 8 | 0.4085 | 0.7442 | 0.2388 | 0.3616 | 0.8703 |
| No log | 3.0 | 12 | 0.3942 | 0.7273 | 0.2388 | 0.3596 | 0.8686 |
| No log | 4.0 | 16 | 0.3650 | 0.6712 | 0.3657 | 0.4734 | 0.8828 |
| No log | 5.0 | 20 | 0.4018 | 0.7234 | 0.2537 | 0.3757 | 0.8739 |
| No log | 6.0 | 24 | 0.3835 | 0.7188 | 0.3433 | 0.4646 | 0.8881 |
| No log | 7.0 | 28 | 0.3521 | 0.6277 | 0.4403 | 0.5175 | 0.8952 |
| No log | 8.0 | 32 | 0.3612 | 0.6047 | 0.3881 | 0.4727 | 0.8917 |
| No log | 9.0 | 36 | 0.3660 | 0.6118 | 0.3881 | 0.4749 | 0.8917 |
| No log | 10.0 | 40 | 0.3585 | 0.5660 | 0.4478 | 0.5 | 0.8917 |
| No log | 11.0 | 44 | 0.3781 | 0.5745 | 0.4030 | 0.4737 | 0.8917 |
| No log | 12.0 | 48 | 0.3847 | 0.5816 | 0.4254 | 0.4914 | 0.8934 |
| No log | 13.0 | 52 | 0.3799 | 0.5962 | 0.4627 | 0.5210 | 0.8934 |
| No log | 14.0 | 56 | 0.3804 | 0.6038 | 0.4776 | 0.5333 | 0.8952 |
| No log | 15.0 | 60 | 0.3894 | 0.6154 | 0.4776 | 0.5378 | 0.8970 |
| No log | 16.0 | 64 | 0.3841 | 0.6038 | 0.4776 | 0.5333 | 0.8952 |
| No log | 17.0 | 68 | 0.3876 | 0.6095 | 0.4776 | 0.5356 | 0.8988 |
| No log | 18.0 | 72 | 0.3945 | 0.6154 | 0.4776 | 0.5378 | 0.8970 |
| No log | 19.0 | 76 | 0.4000 | 0.6275 | 0.4776 | 0.5424 | 0.8988 |
| No log | 20.0 | 80 | 0.4018 | 0.63 | 0.4701 | 0.5385 | 0.8988 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.13.0+cu116
- Datasets 2.8.0
- Tokenizers 0.13.2
