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
To run the model ANMM and DRMM , we need to generate hist and binsum txt file.
But I found in the process.py, the function cal_hist and cal_binsum have some errors. The question and answer matching matrix should be calculated as the cosine similarity value instead of the dot product directly.
The text was updated successfully, but these errors were encountered:
To run the model ANMM and DRMM , we need to generate hist and binsum txt file.
But I found in the process.py, the function cal_hist and cal_binsum have some errors. The question and answer matching matrix should be calculated as the cosine similarity value instead of the dot product directly.
The text was updated successfully, but these errors were encountered: