strisunshinewentingwang/pytorch

Pull Requests Commits

Do not test batched forward grad for inplace ops (#69558)

soulitzer committed 2 years ago

db32daf4

Revert D32606547: torch/monitor: add C++ events and handlers

suo committed 2 years ago

f565167f

[quant][fx][graphmode] Move more patterns to use ModuleReLU fuse handler (#69644)

jerryzh168 committed 2 years ago

f5751799

torch/monitor: add C++ events and handlers (#68783)

d4l3k committed 2 years ago

e61fc1c0

Populate runtime with upgrader graph (#68773)

tugsbayasgalan committed 2 years ago

20f7c893

Back out "[pytorch][PR] Add ability for a mobile::Module to save as flatbuffer" (#69796)

gmagogsfm committed 2 years ago

17f3179d

clear predict_net field from PredictorExporterMeta stored in the exporter to save memory (#68485)

smacke committed 2 years ago

3906f824

[PyTorch][kineto] Remove heap-allocated vectors in saveExtraArgs (#69737)

swolchok committed 2 years ago

19fecc63

Fix the TorchBench CI when running with a benchmark branch. (#69795)

xuzhao9 committed 2 years ago

731c8255

Make c10 tests compilable with -Werror (#69711)

malfet committed 2 years ago

59deee83

Suppress common warnings when building by clang (#69710)

malfet committed 2 years ago

e305e4d4

Revert D32739976: fix CompositeImplicitAutograd ops improperly labeled

malfet committed 2 years ago

41c344d4

Fix docker builds for Python-3.6 (#69785)

malfet committed 2 years ago

77213fa4

[DataPipe] Adding interface for MapDataPipes (#69648)

NivekT committed 2 years ago

a5a7e309

[DataPipe] Adding output types to DataPipe interface file (#69647)

NivekT committed 2 years ago

81a60b98

[PyTorch] Update SmallVector from LLVM (#69110)

swolchok committed 2 years ago

d026057b

Ensure different output filenames for test/train results (#69784)

wconstab committed 2 years ago

Verified 7ec7a37f

[PyTorch] Simple refcount bump fixes in standardizeVectorForUnion & callees (#66695)

swolchok committed 2 years ago

1d269e8c

Move internal ops under ts_backend

desertfire committed 2 years ago

f3229d64

Upstream several internal ops used in LazyTensor, Part 2

desertfire committed 2 years ago

d6224f24

Upstream several internal ops used in LazyTensor

desertfire committed 2 years ago

dd3e15a3

[shard] fix with_comms wrapper (#69493)

wanchaol committed 2 years ago

5374d5d8

[JIT] simplify logic for merging types during profiling (#69096)

davidberard98 committed 2 years ago

e1c583a6

Make vec512 bfloat16 map function clang-Wall clean (#69707)

malfet committed 2 years ago

3219f6a4

Cleanup ProcessGroup.cpp (#69706)

malfet committed 2 years ago

a5ad2cda

Make blend operations clang-Wall clean (#69705)

malfet committed 2 years ago

7ea59261

fix CompositeImplicitAutograd ops improperly labeled (#69169)

bdhirsh committed 2 years ago

195b0d06

Clean up lazy_bench so it runs working models only by default (#69749)

wconstab committed 2 years ago

Verified 1b04924e

use irange for loops 2 (#66746)

r-barnes committed 2 years ago

29d75994

[Jit] Fix schema of aten::split int[] version (#69745)

Hao Lu committed 2 years ago

91d16cb6

Newer Older