mirror of
https://github.com/saymrwulf/pytorch.git
synced 2026-05-15 21:00:47 +00:00
945d7a7408
Summary: As described in https://github.com/pytorch/pytorch/issues/32345, a prototype implementation to add an alltoall communication primitive to torch.distributed module and ProcessGroup abstract interface. Also, implements alltoall in ProcessGroupMPI backend. mnaumovfb JianpingChen066 dmudiger srinivas212 Jianhui-Li mshiryaev ftian1 cc pietern mrshenli pritamdamania87 zhaojuanmao satgera rohan-varma gqchen aazzolini xush6528 osalpekar Pull Request resolved: https://github.com/pytorch/pytorch/pull/32361 Reviewed By: mrshenli Differential Revision: D20635481 Pulled By: srinivas212 fbshipit-source-id: 3dd0af800ce55d02f02813cde550e3a0f1a287d2 |
||
|---|---|---|
| .. | ||
| _static | ||
| _templates | ||
| _templates-stable | ||
| community | ||
| notes | ||
| org/pytorch | ||
| rpc | ||
| scripts | ||
| __config__.rst | ||
| amp.rst | ||
| autograd.rst | ||
| bottleneck.rst | ||
| checkpoint.rst | ||
| conf.py | ||
| cpp_extension.rst | ||
| cuda.rst | ||
| cuda_deterministic.rst | ||
| cuda_deterministic_backward.rst | ||
| cudnn_deterministic.rst | ||
| cudnn_persistent_rnn.rst | ||
| data.rst | ||
| distributed.rst | ||
| distributions.rst | ||
| dlpack.rst | ||
| hub.rst | ||
| index.rst | ||
| jit.rst | ||
| jit_builtin_functions.rst | ||
| jit_language_reference.rst | ||
| jit_python_reference.rst | ||
| jit_unsupported.rst | ||
| math-quantizer-equation.png | ||
| model_zoo.rst | ||
| multiprocessing.rst | ||
| name_inference.rst | ||
| named_tensor.rst | ||
| nn.functional.rst | ||
| nn.init.rst | ||
| nn.rst | ||
| onnx.rst | ||
| optim.rst | ||
| packages.rst | ||
| quantization.rst | ||
| random.rst | ||
| sparse.rst | ||
| storage.rst | ||
| tensor_attributes.rst | ||
| tensor_view.rst | ||
| tensorboard.rst | ||
| tensors.rst | ||
| torch.rst | ||
| type_info.rst | ||
