pytorch/torch/utils/data
Vitaly Fedyunin e9b3bc2ead [DataLoader] Locking lower ranks seed recepients (#81071)
Exit seed receiving section only when all ranks received seed, otherwise we are at risk that current rank
will reach same section of the code again while rank zero still in the previous iteration

Fixes: #80845

Differential Revision: [D37702557](https://our.internmc.facebook.com/intern/diff/D37702557)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/81071
Approved by: https://github.com/msaroufim, https://github.com/ejguan
2022-07-08 18:53:45 +00:00
..
_utils Fix distributed store to use add for the counter of DL shared seed (#80348) 2022-06-27 21:59:17 +00:00
communication
datapipes Expanding DataPipe to support DataFrames (#71931) 2022-07-08 18:46:10 +00:00
__init__.py
backward_compatibility.py [DataLoader] DataLoader now automatically apply sharding to DataPipes 2022-06-02 17:40:29 +00:00
dataframes_pipes.ipynb
dataloader.py [DataLoader] Locking lower ranks seed recepients (#81071) 2022-07-08 18:53:45 +00:00
dataloader_experimental.py [DataLoader] DataLoader now automatically apply sharding to DataPipes 2022-06-02 17:40:29 +00:00
dataset.py Support percentages in random_split (#78877) 2022-06-16 02:00:25 +00:00
distributed.py
graph.py
graph_settings.py
sampler.py added shape checking to WeightedRandomSampler (#78585) 2022-06-02 21:12:14 +00:00
standard_pipes.ipynb
typing.ipynb