Open bieyl opened 2 months ago
Hello! The run log of train.py shows that AVG ACC is always close to 0, e.g. 0.0, 0.05, is this normal? Thanks!
Loss.backward() in the train.py file, can't locate the location of the backward function, does this step actually work in the run?
Hello! The run log of train.py shows that AVG ACC is always close to 0, e.g. 0.0, 0.05, is this normal? Thanks!