Top pytorch-lightning issues

No error message when distributed_backend = "invalid choice", Trainer runs on CPU

bug / fix help wanted
kyoungrok0517 picture kyoungrok0517·5 Sep 2020·9 Comments

Incorrect Precision/Recall/F1 score compared to sklearn

Metrics Priority P0 bug / fix help wanted
junwen-austin picture junwen-austin·18 Aug 2020·14 Comments

How to use multiple metric monitors in ModelCheckpoint callback?

API / design Important discussion question
angshine picture angshine·11 Aug 2020·7 Comments
👍1

Support for Scheduler taking a value to the step function

discussion enhancement won't fix
Alicegaz picture Alicegaz·6 Aug 2020·11 Comments

Disabling automatic .train() for loss criteria

question won't fix
eladrich picture eladrich·4 Aug 2020·14 Comments

Gradient clipping: norm is always 2, device might be undefined.

API / design discussion enhancement help wanted won't fix
PhilJd picture PhilJd·31 Jul 2020·8 Comments

TPU MNIST demo hangs in last batch

bug / fix help wanted
hfwittmann picture hfwittmann·29 Jun 2020·22 Comments
👍3

0.8.2 calls backward on '_GeneratorContextManager'

bug / fix help wanted
s-rog picture s-rog·29 Jun 2020·15 Comments

Understanding Metrics + DDP

Metrics question won't fix
wasserth picture wasserth·24 Jun 2020·8 Comments

Tensorboard logging by epoch instead of by step

Logger question
mpaepper picture mpaepper·8 Jun 2020·12 Comments
 Previous issues
Next issues