strisunshinewentingwang/pytorch

Pull Requests Commits

[Static Runtime] Test isOptimizableContainerType and getAlwaysAliveValues (#65849)

Mike Iovine committed 2 years ago

ed50fa25

Compile without -Wno-unused-variable (take 2) (#66041)

malfet committed 2 years ago

4c4525fa

[FX] Support torch.layout as arg (#66048)

Yinghai Lu committed 2 years ago

6b0aa295

[ao_migration] torch.quantization --> torch.ao.quantization in caffe2/torch/fx (#66096)

jerryzh168 committed 2 years ago

6ea4902c

Binary building wthout python fix (#66031)

n-v-k committed 2 years ago

de24faec

Run master clang-tidy on PRs (#66104)

malfet committed 2 years ago

6eb3a1c8

[PyTorch Edge] Avoid string copying in TypeParser (#64278)

swolchok committed 2 years ago

7c758759

GHA: make obvious when we are running smoke tests to user (#66011)

janeyx99 committed 2 years ago

69da4b43

[Reland] Avoid saving self for `softmax` and `log_softmax` (#66018)

soulitzer committed 2 years ago

4cdfcedd

Refactor functional api vectorized jacobian to use batched grad parameter (#65566)

soulitzer committed 2 years ago

8f5631b8

Add batched_grad parameter to `autograd.grad` (#65564)

soulitzer committed 2 years ago

73901b09

Allow None to pass through for vmap (#65565)

soulitzer committed 2 years ago

b6d5f1ee

[Static Runtime] Fix bug of creating output aliases in aten::embedding_bag (#65516)

d1jang committed 2 years ago

89ed9bda

Fix LLVM-12 UB in generate_proposals_op.cc (#66009)

r-barnes committed 2 years ago

40948a93

Added validation of mode parameter in AveragedModel (#65921)

prabhat00155 committed 2 years ago

c7748fc1

[gpu ne eval] disable adam decay unit test for gpu (#66056)

Hector Yuen committed 2 years ago

0fc6bd2e

Back out "[caffe2] fix LLVM-12 nullptr-with-nonzero-offset UBSAN error" (#66055)

mehtanirav committed 2 years ago

29c0725e

[WIP] skip constant folding dequant node (#63991)

Krovatkin committed 2 years ago

7c529633

Merge remote-tracking branch 'upstream/viable/strict' into bincount_opinfo

saketh-are committed 2 years ago

d1c5ac5b

selective trt import based whether we have gpu or not (#66045)

Yinghai Lu committed 2 years ago

8a307640

[PyTorch Edge] Skip writing version during backport (#65842)

cccclai committed 2 years ago

8b8012a1

[JIT] Selectively enable precise alias analysis for TupleConstruct (#66025)

d1jang committed 2 years ago

7941590a

Revert D31326599: [pytorch][PR] Compile without -Wno-unused-variable

malfet committed 2 years ago

e4ee5ca6

Revert D31359010: [pytorch][PR] Fix cang-tidy regressions caused by #65954

malfet committed 2 years ago

5ef350d7

Fix cang-tidy regressions caused by #65954 (#66040)

malfet committed 2 years ago

c269f471

Fix nll_backward for negative weights (#64572)

lezcano committed 2 years ago

ca76e193

[XROS][ML] System specific adjustments for UTs to work. (#65245)

Karol Kosik committed 2 years ago

eb3b9fe7

GELU acc OP (#65957)

FusionPower committed 2 years ago

363ccb25

Compile without -Wno-unused-variable (#65954)

malfet committed 2 years ago

a6280ab6

Fix shape inference dim_type for Clip, Mean, Div (#65996)

qxy11 committed 2 years ago

10f62942

Newer Older