strisunshinewentingwang/pytorch

Pull Requests Commits

[easy] fix nested view call taking in more than one -1 (#86134)

mikaylagawarecki committed 2 years ago

2494c318

MPS: Add multinomial op (#80760)

kulinseth committed 2 years ago

6a842e33

Added _unsafe_view decomp (#86103)

Chillee committed 2 years ago

37013bb4

[MPS] Cast dot inputs to int32 when needed (#86140)

abhudev committed 2 years ago

40a8cc28

Correctly error if you pass in tensors where size arguments expected (#86126)

ezyang committed 2 years ago

954660a3

Symintified all functions, not including factory functions (#86078)

ezyang committed 2 years ago

2aa9e075

Decay integer-only (Optional)SymIntArrayRef to IntList in IValue (#86094)

ezyang committed 2 years ago

cb87983c

Preserve/strip OptionalSymIntArrayRef when finding real schema (#86114)

ezyang committed 2 years ago

146db41e

Fix compile error for vs2022 #79358 (#85958)

waruqi committed 2 years ago

1da74929

[ONNX] Remove registration in __init__ (#86130)

justinchuby committed 2 years ago

36634d78

Re-introduce the functorch docs build (#85838) (#86125)

zou3519 committed 2 years ago

e01d616b

[MPS] Improve memory usage and performance utilizing garbage collector and adaptive commit (#86119)

razarmehr committed 2 years ago

75c0e3a4

[MPS] Handle compatible inputs to where (#85946)

abhudev committed 2 years ago

8860e489

[GHF] Add commit statuses to checkruns conclusions (#86129)

malfet committed 2 years ago

2f692236

Custom sdp implementations dense (#85984)

drisspg committed 2 years ago

cd647761

[skip-ci] Fixed bad link in build_ci_governance.rst (#85933)

vfdev-5 committed 2 years ago

8d9472d7

[docs] Add `torch.channels_last_3d (#85888)

crcrpar committed 2 years ago

85d520d4

[FSDP] Delay moving tensor to CPU until necessary for optim_state_dict() (#85761)

fegin committed 2 years ago

2067b768

Revert "Require bias to be contiguous for depthwise3x3_winograd backend (#85711)"

pytorchmergebot committed 2 years ago

e23cede0

Update dist.scatter() documentation (#86069)

zerolinux5 committed 2 years ago

c670bad7

implementation of qmul using xnnpack (#86040)

Cuiqing Li committed 2 years ago

2403d0c2

parallelize at file granularity (#85770)

clee2000 committed 2 years ago

7941b042

Added roundup_bypass_threshold_mb knobs to the PyTorch Caching Allocator (#85940)

c-odrin committed 2 years ago

d401732b

Unwrap SymInt => Proxy when being returned from the wrapped function make_fx traces (#86098)

Chillee committed 2 years ago

bc993e39

Fix functorch/test/test_control_flow (#85981)

zou3519 committed 2 years ago

470f8fb9

Change torch.autograd.graph.disable_saved_tensors_hooks to be public API (#85994)

zou3519 committed 2 years ago

a262ccea

Disable XLA test (#86123)

huydhn committed 2 years ago

6d06be89

Revert "Re-introduce the functorch docs build (#85838)"

pytorchmergebot committed 2 years ago

5fa84010

Require bias to be contiguous for depthwise3x3_winograd backend (#85711)

soulitzer committed 2 years ago

9a126702

[Metal][BE] Fix signed/unsigned compare (#86068)

malfet committed 2 years ago

d253d6ec

Newer Older