ganslate-team/ganslate

Document + support framework for fine-tuning on a different dataset

surajpaib opened this issue · 2 comments

A potential usecase from Maastro to use the proton trained models on photon data from a different scanner.

@ibro45 Should we start working on this slowly?

should we change the way the checkpoints are saved and loaded? so that it makes this kind of scenario more intuitive