I updated the download_datasets.py. In original implementation, even if I choose dataset of chembl22 or zinc12, I can't direct use preprocess.py (the fields in .h5 is different as expected). Also, the url of chembl22 should be updated from latest to stable release.
I updated the download_datasets.py. In original implementation, even if I choose dataset of chembl22 or zinc12, I can't direct use preprocess.py (the fields in .h5 is different as expected). Also, the url of chembl22 should be updated from latest to stable release.