https://github.com/boostcampaitech5/level2_klue-nlp-03

level2_klue-nlp-03 created by GitHub Classroom

Science Score: 10.0%

This score indicates how likely this project is to be science-related based on various indicators:

○
CITATION.cff file
○
codemeta.json file
○
.zenodo.json file
○
DOI references
✓
Academic publication links
Links to: arxiv.org
○
Academic email domains
○
Institutional organization owner
○
JOSS paper metadata
○
Scientific vocabulary similarity
Low similarity (4.4%) to scientific vocabulary

Last synced: 10 months ago · JSON representation

Repository

level2_klue-nlp-03 created by GitHub Classroom

Basic Info

Host: GitHub
Owner: boostcampaitech5
Language: Jupyter Notebook
Default Branch: master
Size: 1.8 MB

Statistics

Stars: 0
Watchers: 0
Forks: 1
Open Issues: 6
Releases: 0

Created about 3 years ago · Last pushed about 3 years ago

https://github.com/boostcampaitech5/level2_klue-nlp-03/blob/master/

# level2-klue-nlp-03

## Members

||||||
| --- | --- | --- | --- | --- |
| [](https://github.com/torchtorchkimtorch) | [](https://github.com/KimuGenie) | [](https://github.com/tjddn0402) | [](https://github.com/dnjdsxor21) | [](https://github.com/jjsyeon) |

## Relation Extraction 
Relation Extraction is the task of predicting attributes and relations between subject and object entity in sentence. 
![image](https://github.com/boostcampaitech5/level2_klue-nlp-03/assets/75467530/795279d0-c017-4eab-8def-3f5621189882)


## Getting started

```python
# Install Packages
pip install -r requirements.txt

# Train & Inference
python src/train.py
```
  
### Structure

```python
root/
|
|-- config.yaml
|-- config_sweep.yaml
|
|-- src/
|   |-- train.py
|   |-- train_sweep.py
|   |-- loader.py
|   |-- models.py
|   |-- utils.py
|
|-- eda/
|   |-- entity_EDA.ipynb
|   |-- result_eda.ipynb
```

## Data
#### Augmentation
- [Easy Data Augmentation](https://github.com/toriving/KoEDA)
- [Back Translation](https://github.com/ssut/py-googletrans)
#### Preprocessing
- [Hanja Removal](https://github.com/suminb/hanja)
- [Unidecode](https://github.com/avian2/unidecode)

## Model
- [Entity Marker](https://arxiv.org/abs/2102.01373)
- [Binary Classifier](https://www.kaggle.com/code/duongthanhhung/bert-relation-extraction)
- [RECENT(entity relation restriction)](https://arxiv.org/abs/2105.08393)
- [Additional Model for Fine-tuning](https://arxiv.org/abs/1906.03158)