Top pytorch-lightning issues

`num_tpu_cores=8` does not work on kaggle

Priority P0 bug / fix help wanted
lezwon picture lezwon·20 Apr 2020·12 Comments

Output not reduced in DP

DP Priority P0 bug / fix help wanted
Deanamic picture Deanamic·20 Apr 2020·15 Comments

Batch is not split in 'dp' mode when dataloader output is not a tensor

bug / fix help wanted
zlenyk picture zlenyk·16 Apr 2020·4 Comments

Switch from tqdm to fastprogress for jupyter-lab friendly progress bars

enhancement help wanted
tcwalther picture tcwalther·15 Apr 2020·5 Comments

Should the Trainer attributes be protected?

enhancement question won't fix
awaelchli picture awaelchli·15 Apr 2020·10 Comments

ImportError: cannot import name 'WandbLogger' from 'pytorch_lightning.loggers'

Logger bug / fix help wanted
braindotai picture braindotai·14 Apr 2020·7 Comments

EarlyStopping checkpointed state is lagging one epoch behind

bug / fix discussion help wanted
lizhitwo picture lizhitwo·12 Apr 2020·15 Comments

Implement Asynchronous GPU transfer and Training with Multithreading

enhancement help wanted won't fix
HenryJia picture HenryJia·7 Apr 2020·10 Comments
👍1

TPU error: RAM full, page stopped responding and slower than GPU on google colab

TPU bug / fix help wanted information needed
OliverCWY picture OliverCWY·7 Apr 2020·16 Comments
👀2

Loading from best checkpoint

Hacktoberfest enhancement good first issue help wanted won't fix
philip30 picture philip30·6 Apr 2020·8 Comments
👍4 🚀1
 Previous issues
Next issues