Pretrain, finetune ANY AI model of ANY size on 1 or 10,000+ GPUs with zero code changes.
COMMITS
/ tests/trainer/test_data_loading.py March 27, 2022
C
Drop PyTorch 1.7 support (#12432)
Carlos Mocholí committed
March 5, 2022
A
Deprecate `LightningModule.on_pretrain_routine_{start/end}` (#12122)
ananthsub committed
February 21, 2022
C
Remove `Trainer._strategy_type` (#11990)
Carlos Mocholí committed
February 17, 2022
E
fix typos (#11937)
edward-io committed
February 9, 2022
J
Update `tests/trainer/*.py` to use `devices` instead of `gpus` or `ipus` (#11697)
Jv Kyle Eclarin committed
January 6, 2022
K
Rename `_distrib_type` to `_strategy_type` (#11328)
Kaushik B committed
January 5, 2022
D
Deprecate `TrainerDataLoadingMixin` and move logic to `DataConnector` (#11282)
Danielle Pintz committed
December 14, 2021
R
Fix the num_batches value in warning (#10980)
Rohit Gupta committed
December 3, 2021
R
Disable eval dataloaders replacement during overfitting (#10877)
Rohit Gupta committed
November 24, 2021
A
Support re-instantiation for custom DataLoader in Lightning (#10680)
Adrian Wälchli committed
November 19, 2021
A
Extract dataloader utilities from `TrainerDataLoadingMixin` (#10145)
Adrian Wälchli committed
November 16, 2021
C
Skip strategy=ddp_spawn, accelerator=cpu, python>=3.9 tests (#10550)
Carlos Mocholí committed
November 15, 2021
K
Deprecate `DistributedType` in favor of `StrategyType` (#10505)
Kaushik B committed
November 13, 2021
C
Drop torch 1.6 testing (#10390)
Carlos Mocholí committed
November 4, 2021
P
Fix failure when `DataLoader(batch_size=None)` is passed (#10345)
Peter Dudfield committed
C
Update recommendation on `dataloader_idx` (#10318)
Carlos Mocholí committed
October 25, 2021
D
Mark accelerator connector as protected (#10032)
Danielle Pintz committed
October 20, 2021
K
Update strategy flag in docs (#10000)
Kaushik B committed
October 18, 2021
R
Fix logic to check for spawn in worker_check (#9902)
Rohit Gupta committed
October 16, 2021
K
October 12, 2021
A
update tests to not rely on patched dataloaders (#9905)
Adrian Wälchli committed
September 28, 2021
T
[Refactor] Simplify data loading logic around replacing sampler to prevent confusion (#9721)
thomas chaton committed
September 8, 2021
C
Fix `replace_sampler` missing the batch size under specific conditions (#9367)
Carlos Mocholí committed
September 6, 2021
J
CI: precommit - docformatter (#8584)
Jirka Borovec committed
August 16, 2021
J
[Bugfix] Detach Loaders after running entrypoint (#8885)
Justus Schock committed
August 2, 2021
J
CI: yesqa (#8564)
Jirka Borovec committed
July 26, 2021
C
Replace `yapf` with `black` (#7783)
Carlos Mocholí committed
C
Support `DataLoader`s with missing arguments in `replace_sampler` (#8519)
Carlos Mocholí committed
July 2, 2021
A
Support `DDPPlugin` to be used on CPU (#6208)
Adrian Wälchli committed
April 19, 2021
A
Auto-set `DataLoader.worker_init_fn` with `seed_everything` (#6960)
Adrian Wälchli committed
April 9, 2021
R
Update `DataLoader.persistent_workers` warnings in ddp_spawn (#6762)
Roger Shieh committed
March 2, 2021
J
Refactor: runif for spec 6/6 (#6307)
Jirka Borovec committed
J
Refactor: skipif for Windows 2/n (#6268)
Jirka Borovec committed
J
Refactor: skipif for multi - gpus 1/n (#6266)
Jirka Borovec committed
February 9, 2021
J
fix miss-leading imports in tests (#5873)
Jirka Borovec committed
February 6, 2021
J
yapf tests trainer (#5844)
Jirka Borovec committed
February 5, 2021
J
formatting flake8 & isort (#5824)
Jirka Borovec committed
C
[Feat-BugFix] Resolve custom DataLoader (#5745)
chaton committed