According to the original KD-DTI dataset, before processing the data, you should first register a DrugBank account, download the xml dataset and replace the entity id with the entity name in the drugbank.
Then, you can process the data by:
bash preprocess.sh
For more details, please see here.
You can fine-tune the pre-trained BioGPT on the task by:
bash train.sh
We provide our fine-tuned model on the task. See here
You can inference and evalaute the model on the test set by:
bash infer.sh