strisunshinewentingwang/pytorch

Pull Requests Commits

Port silu_backward to structured (#58661)

Freey0 committed 3 years ago

b52849b5

Make some comparisons explicit (#60505)

r-barnes committed 3 years ago

66f01db3

Enhance ProcessGroupWrapper with additional checks + refactor (#60237)

rohan-varma committed 3 years ago

f5341bd5

[torch/distributed] remove outdated FutureWarning in distributed/elastic/util/store.py (#60807)

Kiuk Chung committed 3 years ago

aaea81e3

Paren-matching kernel launch check without external deps (#60778)

r-barnes committed 3 years ago

94cdbbf4

Python error unit tests on delegation of backend_with_compiler_demo (#60689)

Amyh11325 committed 3 years ago

88b0518a

ENH Adds byte support for nll_loss (CUDA) (#60650)

thomasjpfan committed 3 years ago

e63db3ae

Add -I<directory> option to tools/linter/clang_tidy.py (#60745)

1ntEgr8 committed 3 years ago

7f6b2bc2

Don't reference reflection_pad3d in functional.py (#60837)

ngimel committed 3 years ago

5b118a7f

To add Nesterov Adam algorithm for multi-tensor optimizers API (#59165)

iramazanli committed 3 years ago

f0e972a4

[TensorExpr] Add a mechanism to register custom TS->NNC lowerings in TensorExprKernel. (#60804)

Mikhail Zolotukhin committed 3 years ago

3bfe1508

To add Rectified Adam algorithm for multi-tensor optimizers API (#59161)

iramazanli committed 3 years ago

5563f4bd

Support default values on NamedTuple fields (#54682)

Ansley Ussery committed 3 years ago

0fbc471d

fix cuda mem leak check not properly run on master_builds (#60742)

Rong Rong (AI Infra) committed 3 years ago

6b53792f

[Static Runtime] Remove output type constraints (#60669)

Hao Lu committed 3 years ago

e3abccec

Fix missing spaces in error of constant_pad_nd (#60729)

take-cheeze committed 3 years ago

dae25c20

Modernize for-loops in aten (#59598)

r-barnes committed 3 years ago

9a08e87d

supports non-leaf inputs for autograd.backward() function (#60521)

RockingJavaBean committed 3 years ago

7e3a694b

Remove un-used parameter in _trilinear backward (#60673)

albanD committed 3 years ago

056a8e0d

[Model Averaging] Move step out of model averaging API (#60632)

Yi Wang committed 3 years ago

f2622171

Sparse CSR: increase dtype test coverage (#60656)

IvanYashchuk committed 3 years ago

c5f0692b

add channels last for AdapativeMaxPool2d (#48920)

mingfeima committed 3 years ago

dd045ab5

Fix missing #pragma once in jit/method.h

wconstab committed 3 years ago

367aff91

Add CUDA Vital (#58059)

Victor Bittorf committed 3 years ago

8b6487c6

add a boxed CPU fallback kernel (#58065)

bdhirsh committed 3 years ago

9134b0e4

[torch] Module fix on the support of LazyModule on bug #60132 (#60517)

Hongbo Zhang committed 3 years ago

ad69e2fd

faster generate_square_subsequent_mask in nn.Transformer (#60631)

Basil Hosmer committed 3 years ago

cab926b2

Remove `Optional[None]` annotations (#60704)

Ansley Ussery committed 3 years ago

7585783b

Fix doc preview source directory (#60792)

driazati committed 3 years ago

5ed7400b

configurable pre/post LayerNorm in nn.Transformer (#60593)

Basil Hosmer committed 3 years ago

7b933cd9

Newer Older