strisunshinewentingwang/pytorch

Pull Requests Commits

Reland: [TensorExpr] Add methods for inspecting generated code in `TensorExprKernel`. (#57560)

Mikhail Zolotukhin committed 3 years ago

dedaf4fa

Reland: [StaticRuntime] Use NNC's call_raw API to reduce call overheads. (#57553)

Mikhail Zolotukhin committed 3 years ago

9e7814d5

Reland: [TensorExpr] Add `TensorExprKernel::runFast` method. (#57552)

Mikhail Zolotukhin committed 3 years ago

e686c66f

Reland: [TensorExpr] Add `CodeGen::call_raw` method. (#57551)

Mikhail Zolotukhin committed 3 years ago

0bf69278

Relax tp_new so that it is OK to call (#57544)

ezyang committed 3 years ago

da8cc355

Fixed C++ linalg API (#57464)

heitorschueroff committed 3 years ago

c65a1da9

Revert D28197820: [JIT][NNC] add hardswish symbolic gradient and NNC lowering

mrshenli committed 3 years ago

887d0e56

Fix compatibility problem with LSTMs and torch.save (#57558)

zou3519 committed 3 years ago

0787d781

ns for fx: clean up manual string names of related ops (#57210)

vkuzo committed 3 years ago

49adac65

ns for fx: change matching to only match known types (#57186)

vkuzo committed 3 years ago

76f29d53

ns for fx: add more type to relationship mapping (#57184)

vkuzo committed 3 years ago

44bb15cf

ns for fx: move relatedness mapping to mappings file (#57171)

vkuzo committed 3 years ago

a9dc9535

[AutoAccept][Codemod][FBSourceClangFormatLinter] Daily `arc lint --take CLANGFORMAT`

CodemodService FBSourceClangFormatLinterBot committed 3 years ago

9ec68834

mkldnn gelu (#53615)

Krovatkin committed 3 years ago

aeaa91bf

[JIT][NNC] add hardswish symbolic gradient and NNC lowering (#57383)

eellison committed 3 years ago

0142fd0b

Compute nvrtc during libtorch build (#57579)

malfet committed 3 years ago

133d8abb

Update Gloo submodule (#57586)

pbelevich committed 3 years ago

cd9995ae

[codemod] Enforce proper use of emplacy functions

nicovank committed 3 years ago

45a3231b

[RFC] [PyTorch Edge] Simplify error logging in mobile/import.cpp (#55711)

dhruvbird committed 3 years ago

d728491f

Always run as many quick-checks steps as possible (#57572)

samestep committed 3 years ago

eb39da6b

[Dist profiling] Add is_async field (#57253)

rohan-varma committed 3 years ago

7175d491

[nnc][tests] Skip long running tests when using TE interpreter (#57568)

bertmaher committed 3 years ago

151e81b7

fx quant: remove matching hack for binary qhandler (#57470)

vkuzo committed 3 years ago

7c3a30fd

Add futures to ProcessGroupMPI work (but not including Send/Recv) and python DDP comm hook testing (#57214)

agolynski committed 3 years ago

2b6c09c1

.github: Add render_test_results job (#57472)

seemethere committed 3 years ago

8c9e42ba

tools: Add render_junit script (#57327)

seemethere committed 3 years ago

00d6472b

[iOS GPU] [easy] Rename APIs in MPSImageWrapper

xta0 committed 3 years ago

9c547858

Replace windows CUDA 11.2 CI with 11.3 (#57223)

janeyx99 committed 3 years ago

76d9070d

Use proper Google Analytics id (#56578)

NicolasHug committed 3 years ago

1fc89d9f

Implement torch.sort with cub::DeviceSegmentedRadixSort (#56821)

zasdfgbnm committed 3 years ago

383e4510

Newer Older