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
A good starting point is to ponder the differences between train_nih.py (Chest X-Ray 14 dataset) and train_voc.py (Pascal VOC dataset). These files rely on data/nih.py and data/voc.py which you can take a look. This is the right way to do if you are intended to use the same code base.
However, if you are accustomed to training a deep model by yourself. There is nothing stopping you from just utilizing the PYLON model pylon.py with your own training loop. This is the easy way potentially removes a lot of boilerplates.
Activity
konpatp commentedon Aug 31, 2022
A good starting point is to ponder the differences between
train_nih.py(Chest X-Ray 14 dataset) andtrain_voc.py(Pascal VOC dataset). These files rely ondata/nih.pyanddata/voc.pywhich you can take a look. This is the right way to do if you are intended to use the same code base.However, if you are accustomed to training a deep model by yourself. There is nothing stopping you from just utilizing the PYLON model
pylon.pywith your own training loop. This is the easy way potentially removes a lot of boilerplates.