BERT based Text Classification
Use "Text_Classification.ipynb" to run in google colab.
toch 1.6.0
transformers 3.0.2
python 3.6.9
- Data pre-processing
Pre-process the data and split dataset into train and test sets
python process_data.py
- Train Model
Test classification task is based on BERT model. BertForSequenceClassification is used and "dbmdz/bert-base-turkish-cased" is utilized as pre-trained BERT model for Turkish.
python model.py
- Test Model
Saved model is used to test the test dataset.
python test.py