strisunshinewentingwang/pytorch

Pull Requests Commits

Update XLA hash

alanwaketan committed 1 year ago

80e6a7aa

Update XLA hash

alanwaketan committed 1 year ago

cf0bca52

Remove GetExpandDimensions()

alanwaketan committed 1 year ago

849b89e9

[LTC] Remove view

alanwaketan committed 1 year ago

a689434d

SymIntify _copy functionalization kernels (and _copy_out too) (#88572)

ezyang committed 1 year ago

23a3eb37

fallback for scatter_(scalar) (#88210)

Krovatkin committed 1 year ago

896fa8c5

Publicly expose _LRScheduler to LRScheduler (#88503)

janeyx99 committed 1 year ago

0a69c50a

Upload test stats for inductor workflow (#88535)

huydhn committed 1 year ago

05b9e8ec

[torchdynamo] support torch.autograd._profiler_enabled (#88378)

yuguo68 committed 1 year ago

a3752408

Handle pin_memory in refs.randn (#88473)

SherlockNoMad committed 1 year ago

95d57b54

[nvfuser] skip extremal tests on rocm (#88587)

suo committed 1 year ago

bf49dada

Revert "Use sudo when reset NVIDIA devices (#88531)"

pytorchmergebot committed 1 year ago

7bf9db81

Revert "[fix] allow saving python attr on Tensor and Parameter via torch.save (#81616)"

pytorchmergebot committed 1 year ago

78a0ca29

[exir][fx] PassManager error handling (#88520)

angelayi committed 1 year ago

91a40398

[Dynamo] Fix bug: GradMode doesn't carry grad state correctly after graph break (#88537)

yanboliang committed 1 year ago

bd1ffc65

[2/n] Thread PG: add class _World to distributed_c10d.py (#781) (#88471)

Rodrigo Kumpera committed 1 year ago

6663ae55

Clarify rules for which commit is used in CI (#88425)

ZainRizvi committed 1 year ago

fc8f2f66

Upgrade Linux NVIDIA driver to the latest prod version (#88517)

huydhn committed 1 year ago

c407a7b2

Use sudo when reset NVIDIA devices (#88531)

huydhn committed 1 year ago

505486ce

allow XLA folks update the pin (#88527)

Krovatkin committed 1 year ago

cec4bd99

reland "fix as_strided_scatter_backward (#87646)" (#88342)

bdhirsh committed 1 year ago

a16ced03

[Static Runtime] Fix tensor_split sections overload (#88113)

Mike Iovine committed 1 year ago

dd43903f

[xla hash update] update the pinned xla hash (#88070)

pytorchmergebot committed 1 year ago

7076a648

Support sign for HF models like ElectraForQuestionAnswering (#88160)

EikanWang committed 1 year ago

ad27d762

Support reduction vectorization (#87356)

EikanWang committed 1 year ago

a9d37ce8

Explicit vectorization support for TorchInductor (#87068)

EikanWang committed 1 year ago

6541e51f

use faster cache flush in triton benchmarking (#88557)

ngimel committed 1 year ago

a95419b4

[Dynamo] fix torchdynamo's TVM meta schedule backend (#88249)

YJ Shi committed 1 year ago

eda247ee

[inductor] Add lowering for as_strided_scatter (#88379)

peterbell10 committed 1 year ago

791d9ee2

Revert "Reenable optimizer overlap tests (#88439)"

pytorchmergebot committed 1 year ago

81042d3a

Older