strisunshinewentingwang/pytorch

Pull Requests Commits

[PyTorch] Add torch::jit::toString(const Type&) (#66689)

swolchok committed 2 years ago

a3ca4c83

Clean up dead code (#69296)

wconstab committed 2 years ago

855365e9

CUDACachingAllocator: make an error message more accurate. (#69174)

nelhage committed 2 years ago

a813ddf5

Update the documentation for AMP with DataParallel (#69218)

eliobecoolet8 committed 2 years ago

088a4feb

Limit uploading JSONs to trunk (#69385)

janeyx99 committed 2 years ago

80a67cd3

add HPU case for backend_to_string function (#69225)

nelyahu committed 2 years ago

a20b9f8d

[SR] Use std::vector::reserve in GetLivenessMap (#68884)

donaldong committed 2 years ago

6f7a5ddf

Fixed type checking errors in node.py (#68124)

Onyiee committed 2 years ago

ae112645

[DataPipe] Adding ShufflerMapDataPipe (#68606)

NivekT committed 2 years ago

6baaec30

[PyTorch][JIT] Use stack.pop_back() instead of pop(stack) for DROP (#69326)

swolchok committed 2 years ago

3e457395

[PyTorch] Use toObjectRef in JIT interpreter (#69324)

swolchok committed 2 years ago

2c84b010

.github: Propagate WITH_PUSH to docs jobs (#69372)

seemethere committed 2 years ago

5a480831

Expose is_metal_available in header (#68942)

mattjgalloway committed 2 years ago

8f8524a4

Remove columns and ones from slow2d transpose signatures (#68898)

jbschlosser committed 2 years ago

77ca153d

Remove finput and fgrad_input from slow3d signatures (#68897)

jbschlosser committed 2 years ago

7ca2da14

.github: Add credentials for macos test jobs (#69371)

seemethere committed 2 years ago

73d2ca20

[SR][Code cleanup] Typedef/default for kwargs (#69164)

Mike Iovine committed 2 years ago

6ed73544

Revert D32786909: [C10D] [Easy] Use pinned memory for HtoD copies in Reducer:: sync_bucket_indices

janeyx99 committed 2 years ago

b7611724

Revert of adding windows CUDA 11.5 workflow (#69365)

atalman committed 2 years ago

e0fb228e

CMake: Update precompiled header and fix support (#67851)

peterbell10 committed 2 years ago

21919be9

[SR] Factor logic that determines managed tensors out of MemoryPlanner (#68295)

Mike Iovine committed 2 years ago

cc46dc45

[Pytorch Edge] Make Tracer support xirp metal segmentation (#69328)

JacobSzwejbka committed 2 years ago

276cb8f5

Add OpInfos for combinations, cartesian_prod, sum_to_size, ldexp, as_strided (#68853)

saketh-are committed 2 years ago

a07ffe8d

Back out "Add efficient zero tensors" (#69327)

Mark Richardson committed 2 years ago

834bd313

fix for python 3.10 for gradient opinfo (#68113)

ankitaS11 committed 2 years ago

c572a603

Fix some usages of CUDA_VERSION (#69092)

mcarilli committed 2 years ago

572c3e31

[C10D] [Easy] Use pinned memory for HtoD copies in Reducer:: sync_bucket_indices (#69298)

Andrew Tulloch committed 2 years ago

dbc8d9c9

Add operator div (#68528)

Shirong Wu committed 2 years ago

e2c7bd08

Add support for C++ frontend wrapper on Linux (#69094)

malfet committed 2 years ago

bede18b0

THPStorage: Prefer intrusive_ptr over owning raw pointers (#69248)

peterbell10 committed 2 years ago

33c3c539

Newer Older