strisunshinewentingwang/pytorch

Pull Requests Commits

Move functional collectives implementation to python. (#98595)

Rodrigo Kumpera committed 1 year ago

24d90015

grab bag of changes (#98572)

Chillee committed 1 year ago

c75dd7c4

Remove MERGE_IN_PROGRESS when exiting merge (#98611)

malfet committed 1 year ago

9667f261

Revert "[experiment] More procs in CI (#98098)"

pytorchmergebot committed 1 year ago

55724a5e

[CI] Mark vision_maskrcnn as NONDETERMINISTIC (#98570)

desertfire committed 1 year ago

5210d7c4

[quant][pt2e] Add support for a few ops in QNNPackQuantizer to enable quantizing internal model (#98560)

jerryzh168 committed 1 year ago

c5269ad6

Work around CI worker gpu issue for inductor_distributed (#98601)

wconstab committed 1 year ago

89e57744

[pt2] add meta functions for `cummax` and `cummin` (#98552)

nkaretnikov committed 1 year ago

1c226f5a

[Quant] Add get_symmetric_qnnpack_qat_qconfig_mapping (#98569)

mcr229 committed 1 year ago

483fd335

Clean up compile reason logic, report only graph break compiles (#98574)

voznesenskym committed 1 year ago

e016dec6

Add option to log recomps (#98564)

mlazos committed 1 year ago

f55e72c0

[experiment] More procs in CI (#98098)

clee2000 committed 1 year ago

9fd3eba6

Flip Switch Redux (#98341)

eellison committed 1 year ago

e302f083

Don't use f-strings in logging calls (1/X) (#98591)

ezyang committed 1 year ago

16ec7efa

[better_engineering][multiplatform] Repalce host_info() check with select for default_compiler_flags (#98306)

kkolur76 committed 1 year ago

79e14f8f

Skip nnmodule hook guards by default (#98371)

wconstab committed 1 year ago

390c51bf

[devX] make labels only count their own occurences (#98551)

PaliC committed 1 year ago

46d765c1

Add ephemeral merging label (#98543)

PaliC committed 1 year ago

d06662fb

inductor(CPU): support dynamic shape for onednn fusion path (#97230)

XiaobingSuper committed 1 year ago

d643a00e

[Inductor] Fix bug in lowering.slice_ when negative start out of range (#98517)

yanboliang committed 1 year ago

77d9742c

Revert "Reduce includes of CUDACachingAllocator.h (#97072)"

pytorchmergebot committed 1 year ago

45a2f6b7

Reduce overhead in CUDAGraph Trees (#98529)

eellison committed 1 year ago

5c8fea56

[quant][pt2e] QNNPackQuantizer support annotation for resnet18 (#98507)

jerryzh168 committed 1 year ago

616f50da

refactor(add privateuseone floder in aten/src/ATen): add a PrivateUse… (#98127)

ykddd committed 1 year ago

5a537e29

[pt2][inductor] hardcode autotuning names (#98351)

nmacchioni committed 1 year ago

29608fd2

[vision hash update] update the pinned vision hash (#98367)

pytorchmergebot committed 1 year ago

3d8ead7e

Fix off-by-1 error in dynamo coverage stats (#98558)

jansel committed 1 year ago

1fb8428d

Disable test_torchinductor_dynamic_shapes on ASAN (#98544)

huydhn committed 1 year ago

2161be08

[reland][inductor] Enable CudaWrapperCodeGen for non-AOT mode (#98534)

desertfire committed 1 year ago

152d65ae

Update _linux-test.yml (#98317)

weiwangmeta committed 1 year ago

d4dbdee5

Newer Older