Dataset files preparation

[GuptaVishu2002]

Hi, would it be possible to describe how to prepare negative data or provide neg_train.csv and neg_test.csv (and any other additional files/columns such as test_inchi_ik_dict.pkl or 'NCE') to be able to run the workflow to train the model from scratch (from data_trvate.py)? Thank You.

[sohahassoun]

You can find an example of what the csv files should look like here <https://zenodo.org/records/10929401/files/pos_train.csv>. data_trvate.py doesn't use neg_train.csv and neg_test.csv even though it loads those files. You can use empty csv files as a placeholder. If you're looking to train the model from scratch, I suggest you format your data as the example shown and name those files pos_train.csv and post_test.csv, and use empty neg_train.csv and neg_test.csv files. *Soha HassounProfessorDepartment of Computer ScienceTufts University* *http://www.cs.tufts.edu/~soha* <http://www.cs.tufts.edu/~soha>

…

On Sun, Jan 5, 2025 at 7:58 PM GuptaVishu2002 ***@***.***> wrote: Hi, would it be possible to give a description of how to prepare negative data or provide neg_train.csv and neg_test.csv to be able to run the workflow to train the model from scratch (from data_trvate.py)? — Reply to this email directly, view it on GitHub <#8>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AA6ND2SPTT375IPA3DCLC432JHIKRAVCNFSM6AAAAABUUUTLOOVHI2DSMVQWIX3LMV43ASLTON2WKOZSG43DSNJVGIYTOMY> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[GuptaVishu2002]

Thank you for the instructions and quick response!