strisunshinewentingwang/pytorch

Pull Requests Commits

Add an option to allow passing ignored_parameters to fsdp

zhaojuanmao committed 1 year ago

eeb7bdd9

[CI] Disable rocm distributed tests (#91632)

malfet committed 1 year ago

2965d7e1

[MPS] Implement MPSGenerator to enable manual random seeding (#91348)

razarmehr committed 1 year ago

688e3519

inductor: meta registration for mkldnn ops (#91299)

XiaobingSuper committed 1 year ago

dfb65145

Avoid copies in matmul (#76828)

lezcano committed 1 year ago

8c2e82b4

Revert "Avoid copies in matmul (#76828)"

pytorchmergebot committed 1 year ago

db2a2377

symbolic shapes: add parenthesis around FloorDiv expression (#91554)

nunoplopes committed 1 year ago

2b0abd4c

[MPS] Add bincount support for mps (#91267)

DenisVieriu97 committed 1 year ago

f7939b21

adding test to audit CompositeImplicitAutograd ops that do not have a batching rule (#91367)

srossross committed 1 year ago

cb320482

implemented test and Changed assert to TORCH_CHECK #88808 (#91273)

cnemri committed 1 year ago

6e236553

Revert D42257039: Multisect successfully blamed D42257039 for test or build failures (#91548)

weiwangmeta committed 1 year ago

cce577b3

fix inductor linspace when steps=1 (#91578)

jancervenka committed 1 year ago

fae821c2

Avoid copies in matmul (#76828)

lezcano committed 1 year ago

0c365958

Dispatch the auxiliary frobenius_norm and nuclear_norm to better implementations and deprecate them (#81763)

lezcano committed 1 year ago

12224598

Support indices contiguity per batch and non-contiguous values in sparse compressed tensors (#91243)

pearu committed 1 year ago

b797a242

[MPS] Add suport for casting updatesTensor directly in scatter (#91197)

DenisVieriu97 committed 1 year ago

dbf96164

Deduplicate c10 error and PyTorchError hierarchy (#87855)

shikanime committed 1 year ago

34f2d3e6

[xla hash update] update the pinned xla hash (#91087)

pytorchmergebot committed 1 year ago

2b52db9c

Revert "[cuBLAS] Fix default cuBLAS workspace size and parsing for multiple workspaces (#89027)"

pytorchmergebot committed 1 year ago

39d49dbe

Clang-Tidy: Improve ctors by removing unnecessary copies and initializations (#91538)

Skylion007 committed 1 year ago

77c2a8a1

[cuBLAS] Fix default cuBLAS workspace size and parsing for multiple workspaces (#89027)

eqy committed 1 year ago

b407d98d

Remove _ignored_param_names (#91530)

zhaojuanmao committed 1 year ago

f6136331

[BE] Move internal only non-globbed lists to OSS (#91513)

malfet committed 1 year ago

6cef5948

[cuDNN][cuDNN V8 API] Improve hot path heuristics performance in V8 (#90811)

eqy committed 1 year ago

73436af4

Rename `torchtriton` (#91539)

malfet committed 1 year ago

bc92444b

Bump protobuf from 3.20.1 to 3.20.2 in /.github/requirements (#91540)

dependabot[bot] committed 1 year ago

62713636

Cache conda and pip for IOS CI (#91359)

remidomingues committed 1 year ago

fdbbd20f

switch causal mask for is_causal flag (#91171)

mikekgfb committed 1 year ago

af589b3d

Some CMake and CUDA cleanup given recent update to C++17 (#90599)

cyyever committed 1 year ago

9710ac65

Fix NumPy broadcasting in lstsq_backward (#91460)

lezcano committed 1 year ago

d5163f52

Older