Top pytorch-lightning issues

MLFlowLogger throws a JSONDecodeError

bug / fix help wanted information needed
srib picture srib·19 Aug 2020·3 Comments

RuntimeError: No `loss` value in the dictionary returned from `model.training_step()` with pytorch lightning

question
GuptaVishu2002 picture GuptaVishu2002·18 Aug 2020·6 Comments

Results gathering with varying tensor shapes (e.g. last batch)

bug / fix help wanted
awaelchli picture awaelchli·17 Aug 2020·1 Comments
👍1

AttributeError: module 'pytorch_lightning' has no attribute 'TrainResult'

bug / fix help wanted
francoisruty picture francoisruty·14 Aug 2020·7 Comments

Custom Checkpoint callback for multiple models

question
kowshikthopalli picture kowshikthopalli·13 Aug 2020·3 Comments

When the parameter gpus of Trainer> 1, _pickle.PicklingError:

question
guangmingjian picture guangmingjian·8 Aug 2020·10 Comments

This is really a bad idea. what if someone really don't want to sample/shuffle their data during training.

documentation question
chu-NMSU picture chu-NMSU·7 Aug 2020·3 Comments

Cannot pickle custom metric with DDP mode.

Metrics bug / fix help wanted
c00k1ez picture c00k1ez·6 Aug 2020·4 Comments

Do multiple optimizer steps in one training step

question
paudom picture paudom·4 Aug 2020·13 Comments

NumpyMetric not mapping back to GPU in multi-GPU training

bug / fix help wanted
jcreinhold picture jcreinhold·3 Aug 2020·3 Comments
 Previous issues
Next issues