<!-- 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. -->
scibert_scivocab_uncased-finetuned-ner
This model is a fine-tuned version of allenai/scibert_scivocab_uncased on the plo_dunfiltered_config dataset. It achieves the following results on the evaluation set:
- Loss: 0.1390
- Precision: 0.9649
- Recall: 0.9612
- F1: 0.9631
- Accuracy: 0.9594
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: 1e-05
- train_batch_size: 32
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 11
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.1176 | 1.4 | 5000 | 0.1243 | 0.9570 | 0.9511 | 0.9540 | 0.9502 |
| 0.0973 | 2.81 | 10000 | 0.1129 | 0.9609 | 0.9572 | 0.9590 | 0.9553 |
| 0.0721 | 4.21 | 15000 | 0.1198 | 0.9645 | 0.9585 | 0.9615 | 0.9578 |
| 0.0634 | 5.62 | 20000 | 0.1259 | 0.9649 | 0.9589 | 0.9619 | 0.9582 |
| 0.0572 | 7.02 | 25000 | 0.1321 | 0.9653 | 0.9609 | 0.9631 | 0.9594 |
| 0.0472 | 8.43 | 30000 | 0.1390 | 0.9649 | 0.9612 | 0.9631 | 0.9594 |
| 0.0434 | 9.83 | 35000 | 0.1442 | 0.9656 | 0.9613 | 0.9634 | 0.9598 |
Framework versions
- Transformers 4.18.0
- Pytorch 1.10.1+cu111
- Datasets 2.1.0
- Tokenizers 0.12.1
