check norm as class not instance & reset before every epoch instead before every fit

fastai / course22p2

course.fast.ai 2022 part 2

https://course.fast.ai/Lessons/part2.html

Apache License 2.0

467 stars 252 forks source link

Open anubhavmaity opened 5 months ago

anubhavmaity commented 5 months ago

Two changes

Currently we are passing nn.BatchNorm as class not instance, therefore isinstance will not work. We have to compare nn.BatchNorm as class
Currently we reset losses, accuracy, and ns before each fit. It should be done at the start of every epoch instead.

review-notebook-app[bot] commented 5 months ago

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.