Pretrain, finetune ANY AI model of ANY size on 1 or 10,000+ GPUs with zero code changes.
COMMITS
June 30, 2022
S
adding LAI test (#13321)
Sherin Thomas committed
G
fix mypy typing errors in lightning/trainer/optimizers.py (#13470)
Gautier Dagan committed
W
More clear docs for `LightningDataModule` (#13464)
Wang Ran (汪然) committed
W
Typo in trainer/supporters.py (#13455)
Wang Ran (汪然) committed
W
Remove unused argument `model` (#13454)
Wang Ran (汪然) committed
W
Typo in tuner/lr_finder.py (#13453)
Wang Ran (汪然) committed
W
Fix typo in `Loop.replace` docstring (#13452)
Wang Ran (汪然) committed
W
Fix typo in `_block_parallel_sync_behavior` docstring (#13451)
Wang Ran (汪然) committed
J
CI: abstract and make full pkg check (#13460)
Jirka Borovec committed
O
Add BaseModelCheckpoint class to inherit from (#13024)
otaj committed
C
Simplify list extension (#13435)
Carlos Mocholí committed
S
fixed doc of timer (#13393)
S. Kumano committed
E
Add flash[image] dependency in Active learning example (#13442)
Ekagra Ranjan committed
D
Update wandb requirement from <0.12.19,>=0.8.21 to >=0.8.21,<0.12.20 in /requirements (#13415)
dependabot[bot] committed
W
Remove unused docstring parameter `device` (#13448)
Wang Ran (汪然) committed
S
Set timeout for DDPSpawnStrategy (#13383)
Siyuan Li committed
June 29, 2022
A
Fix validation when accelerator is a string (#13417)
Adrian Wälchli committed
R
Convert validation loop config warnings to `PossibleUserWarning` (#13377)
Rhys Goodall committed
D
Update comet-ml requirement from <=3.28.2,>=3.1.12 to >=3.1.12,<3.31.6 in /requirements (#13414)
dependabot[bot] committed
A
Call `set_epoch` for distributed batch samplers (#13396)
Adrian Wälchli committed
A
Remove remaining old-style AcceleratorConnector properties (#13412)
Adrian Wälchli committed
A
Unpin `protobuf` version and update `tensorboard` version (#13259)
Akihiro Nitta committed
W
Fix docstring typo (#13447)
Wang Ran (汪然) committed
J
fix PL release docker (#13439)
Jirka Borovec committed
M
Modified python version check to accommodate for legacy version styles (#13420)
Martino Sorbaro committed
A
Update gather_all_tensors to handle tensors of different sizes (#12630)
ananthsub committed
K
[Docs] Fix README.md in lightning/examples/pl_basics (#13380)
Keiichi Kuroyanagi committed
A
Remove support for DDP2 strategy (#12705)
Adrian Wälchli committed
S
Add model summary when using DeepSpeed Stage 3 (#13427)
Sean Naren committed
J
CI: fix requirements freeze (#13441)
Jirka Borovec committed