strisunshinewentingwang/pytorch

Pull Requests Commits

clee2000 committed 2 years ago

0eca2214

[quant][pt2e][refactor] Cleanup the logic for deciding whether to insert observer/fq or not (#99220)

jerryzh168 committed 2 years ago

d56adb1b

Update docstring of torch.nn.functional.normalize() (#99512)

mega-optimus committed 2 years ago

06081ac8

Delete tracing_mode argument to export (#99555)

ezyang committed 2 years ago

e9786149

Move more stuff to after_aot (#99557)

ezyang committed 2 years ago

881c5723

Do not assume static by default when exporting (#99554)

ezyang committed 2 years ago

d3bb762f

[BE] Build PyTorch with `-Wnewline-eof` (#99687)

malfet committed 2 years ago

6b8ef8ea

Fix torch.nn.FractionalMaxPool2d output_size error (#99507)

tfsingh committed 2 years ago

dbf0db95

Revert "[c10d] Faster coalescing (#98793)"

pytorchmergebot committed 2 years ago

9861ec97

Revert "fix onednn ConvTranspose2d channels last issue when ic=1 (#99539)"

pytorchmergebot committed 2 years ago

da57d597

Init comm_nonblocking_ when creating AutoNcclGroup (#99679)

kwen2501 committed 2 years ago

9df8b1b5

Support record_stream in dispatch mode (#99529)

lw committed 2 years ago

24bf15fe

Pass locals to enum_repr to correctly make the guard str for enums (#99680)

voznesenskym committed 2 years ago

0ac0d9d2

Bug - check config for dynamic (#99676)

voznesenskym committed 2 years ago

8ee59280

[exir][delegate] torch.ops.call_delegate (#92562)

angelayi committed 2 years ago

a421c547

[combined tracebacks] missing gil acquire (#99685)

zdevito committed 2 years ago

9def7990

[inductor] skip triton.Config that spills (#99385)

shunting314 committed 2 years ago

daff0408

Disable XProtect on MacOS runner (#99506)

huydhn committed 2 years ago

9bece55a

Make CI error on inductor fallback when decomp is available (#99473)

wconstab committed 2 years ago

63690afc

[Inductor][quant]Enable decomposed.quant/dequant lowering and vec code gen (#99131)

leslie-fang-intel committed 2 years ago

deaf983b

inductor: fix onednn conv2d(transpose) packed issue when input size is three (#99601)

XiaobingSuper committed 2 years ago

2a47f685

[ONNX] Fix missing import numpy for docs example (#99663)

BowenBao committed 2 years ago

51742a46

Correct typo for NCCL_MAJOR (#99482)

ajindal1 committed 2 years ago

16a4dc0f

Allow in graph einops operators (#99631)

wconstab committed 2 years ago

6427b849

Make testing._internal.common_utils safe to import (#99659)

wconstab committed 2 years ago

716ba685

[dynamo] Fix example_inputs with unsqueeze_ (#98696)

jansel committed 2 years ago

d168161c

[DTensor] Change Sharding algorithm to be in line with ``torch.chunk()`` (#98722)

wz337 committed 2 years ago

0d2b55c4

add storage serialization methods for privateuse1 (#98920)

XDaoHong committed 2 years ago

27f8eb8c

[inductor] Update triton pin (#99209)

peterbell10 committed 2 years ago

907f2dad

Disable SDPA FlashAttention backward and mem eff attention on sm86+ for head_dim above 64 (#99105)

cpuhrsch committed 2 years ago

fdeee436

Older