Downloads · 30 days
11
2% of all-time downloads
biobert-finetuned-ncbi
Umesh/biobert-finetuned-ncbi
biobert-finetuned-ncbi is a token classification model from Umesh. Use it when you need labels on individual words, such as names. It is set up for transformers.
should probably proofread and complete it, then remove this comment. --
All-time downloads
488
Public
Repo size
3.4 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin431 MB · 100%
At a glance
- Task
- Token Classification
- Library
- transformers
- Model type
- bert
- Access
- Public
- Created
- Jan 17, 2023
- Updated
- Jan 24, 2023
- SHA
- 5c06f144
Try a prompt
Datasets
Model card
From the Hugging Face model README
<!-- 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. -->
biobert-finetuned-ncbi
This model is a fine-tuned version of dmis-lab/biobert-v1.1 on the ncbi_disease dataset. It achieves the following results on the evaluation set:
- Loss: 0.0590
- Precision: 0.8193
- Recall: 0.8640
- F1: 0.8411
- Accuracy: 0.9856
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: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.1049 | 1.0 | 680 | 0.0588 | 0.7826 | 0.7776 | 0.7801 | 0.9806 |
| 0.0362 | 2.0 | 1360 | 0.0539 | 0.8084 | 0.8577 | 0.8323 | 0.9852 |
| 0.0109 | 3.0 | 2040 | 0.0590 | 0.8193 | 0.8640 | 0.8411 | 0.9856 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.13.1+cu116
- Datasets 2.8.0
- Tokenizers 0.13.2
- Task
- Token Classification
- Library
- transformers
- Type
- bert
- Created
- Jan 17, 2023
- Updated
- Jan 24, 2023