strisunshinewentingwang/pytorch

Pull Requests Commits

Dispatch for xpu in adaptive_avg_pooling (#84541)

cchheennhhaaoo committed 2 years ago

368742d0

[functorch] vmapvjpvjp (re-enable test with skips and xfails) (#83999)

kshitij12345 committed 2 years ago

eddc2370

Revert "[functorch] vmapvjpvjp (re-enable test with skips and xfails) (#83999)"

pytorchmergebot committed 2 years ago

76fc6905

[functorch] vmapvjpvjp (re-enable test with skips and xfails) (#83999)

kshitij12345 committed 2 years ago

9addeccb

Changes to prepare for fake tensors on in functorch by default (#84432)

eellison committed 2 years ago

8bd9fe3f

[Profiler] Add quoted metadata API to remove empty trace cpu_op metadata (#84128)

aaronenyeshi committed 2 years ago

b288cfd3

[ONNX] Align Optional Type in block (#83599)

titaiwangms committed 2 years ago

942c0f31

Suggest draft PRs in contribution_guide.rst (#84658)

kit1980 committed 2 years ago

49ec8d32

Preserver stacktrace over functionalization (#84662)

SherlockNoMad committed 2 years ago

0945074a

[vision hash update] update the pinned vision hash (#84679)

pytorchmergebot committed 2 years ago

cb6ba27d

[torchdynamo hash update] update the pinned torchdynamo hash (#84678)

pytorchmergebot committed 2 years ago

889540d0

Remove caffe2 mobile (#84338)

jmdetloff committed 2 years ago

e0229d65

Ignore UB on multiply (#84665)

ezyang committed 2 years ago

9669e3c6

Actually chown artifacts (#84672)

malfet committed 2 years ago

1a1bcc73

Convert ConcretePyInterpreterVTable into Meyer singleton (#84657)

ezyang committed 2 years ago

93359bf9

Convert NoopPyInterpreterVTable into a Meyer singleton (#84656)

ezyang committed 2 years ago

9162bc02

[functorch] add pinv batch rule (#83761)

samdow committed 2 years ago

29672b21

Add underlying_store property for PrefixStore (#84640)

H-Huang committed 2 years ago

586832ce

Turn on functionalization by default in functorch (#84435)

eellison committed 2 years ago

e68df8e4

check rate limits of other tokens too (#83632)

clee2000 committed 2 years ago

6b211161

[functorch] add matrix_rank rule (#83760)

samdow committed 2 years ago

9532c7e2

Add host and port to TCPStore pyi definition (#84636)

H-Huang committed 2 years ago

e14f46f9

[PyTorch] Guard against self assignment in SymInt (#84375)

swolchok committed 2 years ago

c7f6deb6

Fix error in the index range math expression in the docstring of MultiMarginLoss (#84513)

wenh06 committed 2 years ago

fc4acd44

[CUBLAS][TF32][CUDNN] Update numerical_accuracy.rst (#79537)

eqy committed 2 years ago

d892d5d6

Revert "Call jit decomposition in VariableType to increase forward AD coverage (#84151)"

pytorchmergebot committed 2 years ago

acb4a096

add option to remove passes (#84425)

Wei Wei committed 2 years ago

31ef8ddb

Improve torch::jit::as_{module,object} performance (#84399)

peterbell10 committed 2 years ago

2feb31cb

Add CUDA Sanitizer (#83984)

sypneiwski committed 2 years ago

2b2e0fdd

Make dispatcher registrations of SymInt functions backwards compatible (#84557)

ezyang committed 2 years ago

19e27b15

Newer Older