|
--- |
|
license: apache-2.0 |
|
base_model: bert-base-uncased |
|
tags: |
|
- 'biology ' |
|
- NLP |
|
- text-classification |
|
- drugs |
|
- BERT |
|
metrics: |
|
- accuracy |
|
- precision |
|
- recall |
|
- f1 |
|
model-index: |
|
- name: bert-drug-review-to-condition |
|
results: [] |
|
language: |
|
- en |
|
library_name: transformers |
|
datasets: |
|
- Zakia/drugscom_reviews |
|
--- |
|
|
|
# bert-drug-review-to-condition |
|
|
|
This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.4308 |
|
- Accuracy: 0.9209 |
|
- Precision: 0.9061 |
|
- Recall: 0.9209 |
|
- F1: 0.9106 |
|
|
|
## Model description |
|
|
|
Fine-tuning of Bert model with drug-related data for the purpose of text classification |
|
|
|
## Intended uses & limitations |
|
|
|
Personal project. |
|
|
|
## Training and evaluation data |
|
|
|
Kallumadi,Surya and Grer,Felix. (2018). Drug Reviews (Drugs.com). UCI Machine Learning Repository. https://doi.org/10.24432/C5SK5S. |
|
|
|
## Training procedure |
|
Multiclass classification |
|
The model predicts the 'condition' feature from the 'review' feature, only the first 21 conditions are selected. |
|
The 'review' feature is lowercased, we select only values with at least 16 characters. |
|
|
|
### Training hyperparameters |
|
|
|
The following hyperparameters were used during training: |
|
- learning_rate: 5e-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: 3.0 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:| |
|
| No log | 1.0 | 113 | 1.1375 | 0.7747 | 0.7301 | 0.7747 | 0.7450 | |
|
| No log | 2.0 | 226 | 0.5595 | 0.8854 | 0.8675 | 0.8854 | 0.8728 | |
|
| No log | 3.0 | 339 | 0.4308 | 0.9209 | 0.9061 | 0.9209 | 0.9106 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.40.0 |
|
- Pytorch 2.2.1+cu121 |
|
- Datasets 2.19.0 |
|
- Tokenizers 0.19.1 |
