strisunshinewentingwang/pytorch

Pull Requests Commits

Try to fix shutdown test in edge cases

albanD committed 2 years ago

9b1ca6f6

Revert "Cleanup PT-D imports (#85781)"

pytorchmergebot committed 2 years ago

0e639ff4

COO intersection primitives : fusing value selection with value intersection. (#86269)

nikitaved committed 2 years ago

9b2ea41f

[autocast] Clean up registrations using new macros (#86403)

zou3519 committed 2 years ago

e125baf9

[autocast] Make it easier to register rules (#86402)

zou3519 committed 2 years ago

9b74267e

remove unused arg from `impl_func_cum_ops` (#86364)

suprajb committed 2 years ago

55f5e0de

Relandx3 "SymIntify cat and narrow" (#86289)

ezyang committed 2 years ago

a00f8489

Update distributed.rst backend collective support chart (#86406)

H-Huang committed 2 years ago

cc9183eb

[fix] sum_to_size: MathBits test - don't reuse same input tensor (#86378)

kshitij12345 committed 2 years ago

b74ca31b

Override Quantized Backend to use Fbgemm in Qlinear Packed Params Test (#86236)

salilsdesai committed 2 years ago

facbddb9

[Profiler] record gradient from nnModule (#86355)

slgong-fb committed 2 years ago

dbea07b6

Fix col2im and im2col decompositions (#86426)

lezcano committed 2 years ago

28a0b3fb

Improve interpolate() speed for channels_last images and masks (#86361)

NicolasHug committed 2 years ago

93b2d991

Fix the performance issue that the for-loop before ExternallCall could not be parallelized. (#85056)

EikanWang committed 2 years ago

70c6a988

Revert "Revert "Revert "SymIntify cat and narrow (#86191)"" (#86289)"

pytorchmergebot committed 2 years ago

2110c894

[CuDNN v8 API][Quantization]fix alignment function in quantized cuDNN V8 path (#86253)

eqy committed 2 years ago

6c604c92

Introduce a match filter for SubgraphRewriter (#86430)

SherlockNoMad committed 2 years ago

455b8739

[torchdynamo hash update] update the pinned torchdynamo hash (#86399)

pytorchmergebot committed 2 years ago

b5fd845f

[MPS] Cache multinomial_with_replacement graph (#86437)

malfet committed 2 years ago

10aead9a

Fix unfold backward decomp aliasing for 0 dim input (#86428)

eellison committed 2 years ago

9ceadcad

Add Skip List for Aten Ops that are fused in nvFuser. (#86101)

kevinstephano committed 2 years ago

b14f1d7b

Xformer SDP forward/backward kernel (#86157)

drisspg committed 2 years ago

c5a48440

[vision hash update] update the pinned vision hash (#86173)

pytorchmergebot committed 2 years ago

ca39e367

Fix SubgraphMatcher for case of no anchor found (#86421)

SherlockNoMad committed 2 years ago

2fec853c

Fix cond tests after CI was disabled for a bit (#86321)

voznesenskym committed 2 years ago

b73f0e98

Fix broadcasting to implicit leading dimensions in `torch.where` on MPS (#86240)

aslvrstn committed 2 years ago

ca69ddb4

[refactor] Renaming ao.sparsity to ao.pruning (#84867)

z-a-f committed 2 years ago

0e30da3f

Cleanup PT-D imports (#85781)

dstaay-fb committed 2 years ago

9a170b24

[nll_loss] Avoid unnecessary type casts (#86086)

crcrpar committed 2 years ago

a2419638

Replacement is irrelevant for 1-sample multinomial (#86342)

malfet committed 2 years ago

2232db7f

Older