You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Fixed critical bug in smiles_parser.py, where explicit_hydrogens was set to False instead of True, leading to significant degradation of screening power since the released model was not trained to deal with implicit hydrogens.
Added the --batch_size option (defaults to 1), useful for greatly speeding up screening especially when there are CUDA devices specified in --bind_device
Added the --precision option (defaults to 5) which controls the number of decimal places to output