strisunshinewentingwang/pytorch

Pull Requests Commits

Add context manager for conditional rewrites of torch.* to torch._refs.* calls (#81764)

IvanYashchuk committed 2 years ago

900e93d3

[xla hash update] update the pinned xla hash (#82640)

pytorchmergebot committed 2 years ago

54c5f5a5

Abstract away git repo during tests (#82594)

ZainRizvi committed 2 years ago

87a11f11

CoreML backend should log on failure not success (#82604)

jmdetloff committed 2 years ago

7d490dba

Fake: copy over grad attribute (#82593)

eellison committed 2 years ago

09a6d0b5

guard against issues where we try to access tensor_meta on non-tensor input (#82611)

Chillee committed 2 years ago

c6b1e096

Fix: memory cross-border access on the ROCM platform (#76100)

xu.wang committed 2 years ago

62de55ed

Revert "Fix crash on unload torch cpu dll (#67632)"

pytorchmergebot committed 2 years ago

41b54c30

[Profiler] Clean up visit logic (#80822)

Taylor Robie committed 2 years ago

6ada85ef

integrate plugin (#82395)

Shirong Wu committed 2 years ago

09059d91

Fix false positive AVX, AVX2 and AVX512 detection with MSVC (#82554)

JohT committed 2 years ago

2ffb2361

[_shard] add copy_ to shardedtensor (#82508)

wanchaol committed 2 years ago

48a34acf

Add check() to FileCheck type stub. (#82516)

zhxchen17 committed 2 years ago

b8c1aa81

[ROCm] Enable/fix unit tests test_stream_args and test_event_args (#82346)

alugorey committed 2 years ago

f1a13569

Initial private SDP interface and naive composite impl (#81956)

jbschlosser committed 2 years ago

6ca95547

[mergebot] Add Land Check Troubleshooting Message (#82580)

zengk95 committed 2 years ago

c041b2f1

Add tests for fake tensor striding (#82571)

eellison committed 2 years ago

9b46737f

Add tests for aliasing in fake tensor (#82337)

eellison committed 2 years ago

b2f6aa66

Add Autocast Support for FakeTensors / use fake device dispatch keys (#82449)

eellison committed 2 years ago

642aed8b

Fixed QuantizedMeta creation via empty (#82587)

ezyang committed 2 years ago

a64c981d

Fix a NotImplemented mode bug and improve Parameter handling for fake tensor (#82574)

ezyang committed 2 years ago

bf387e89

Correctly unpack constants when used in multi-return output (#82568)

ezyang committed 2 years ago

98215923

Revert "Add a lint rule for torch/csrc/util/pybind.h include (#82552)"

pytorchmergebot committed 2 years ago

532b8a9e

Revert "Add a lint rule for PYBIND11_DECLARE_HOLDER_TYPE (#82556)"

pytorchmergebot committed 2 years ago

ab8e5e6f

[fx] Minor modifications to pass infra (#82485)

angelayi committed 2 years ago

e06d1029

Move doxygen generation to nightly only (#82573)

janeyx99 committed 2 years ago

207ac1c5

[MPS] Fix `torch.full` for boolean types (#82575)

malfet committed 2 years ago

bdd0a4a8

Lintrunner: Run mypy-strict on torchgen (#82576)

peterbell10 committed 2 years ago

afafd166

Store non-symbolic SymInt as int in IValue (#82555)

ezyang committed 2 years ago

9f0fccad

Use OrderedSet in ufunc codegen (#82567)

peterbell10 committed 2 years ago

ba84e966

Newer Older