strisunshinewentingwang/pytorch

Pull Requests Commits

Do rowwisemoments computation in `float` for `half` `LayerNorm` (#66920)

eqy committed 2 years ago

d9c4b3fe

[JIT] Re-enable alias sensitive peepholes (#65860)

Elias Ellison committed 2 years ago

6e6ede2e

[Static Runtime] Bundle function & function_kind to carry them together (#66974)

d1jang committed 2 years ago

051ea5cc

Fix dropout backward by not tracing it (#66823)

wconstab committed 2 years ago

Verified 912530bb

Fix ArchiveReader to keep archive path (#67035)

ejguan committed 2 years ago

3d7a344c

Skip meta conversion before running shape inference (#67056)

bdhirsh committed 2 years ago

Verified bf34713b

Replace lazy_tensors::StrJoin with c10::Join

desertfire committed 2 years ago

4bc9a024

remove accscalar from i0 and i0e (#67048)

ngimel committed 2 years ago

d1a5612a

[PyTorch Edge][type] Add type support for NamedTuple custom class (import) (#63130)

cccclai committed 2 years ago

5f58764d

Implement forward AD for linalg.matrix_exp (#62716)

lezcano committed 2 years ago

d3fc3c4d

diff tool (#66854)

qihqi committed 2 years ago

fe102b98

[quant][fx][graphmode] Rename files and functions for convert and add do_not_use suffix (#66955)

jerryzh168 committed 2 years ago

8ea985f2

[iOS] Bump up iOS CocoaPods version to 1.10.0 (#67058)

Hanton Yang committed 2 years ago

01ced452

Do not build PyTorch with caffe2 by default (#66658)

malfet committed 2 years ago

77beccae

made functorch not decompose by default (#66945)

Chillee committed 2 years ago

4fe8055b

Fixes CUDA vs CPU consistency for index_put_ when accumulating (#66790)

vfdev-5 committed 2 years ago

28fac234

Enroll bowangbj@ to PyTorch distributed package (#67062)

bowangbj committed 2 years ago

35965869

Revert D31838513: Strided masked reduction: mean.

ngimel committed 2 years ago

20f08d23

[skip ci] Set test owner for test_cuda* tests (#66838)

janeyx99 committed 2 years ago

2578de48

Strided masked reduction: mean. (#66784)

pearu committed 2 years ago

b40a9401

Binaries without AVX512 kernels shouldn't report CPU Capability as AVX512 on machines with AVX512 support (#66703)

imaginary-person committed 2 years ago

b696d64e

Implement histogramdd on CPU (#65318)

saketh-are committed 2 years ago

33790c4e

Set test owners for quantization tests (#66832)

janeyx99 committed 2 years ago

6a224b33

Revert D31474901: [pytorch][PR] [numpy] add torch.argwhere

ngimel committed 2 years ago

f29e5220

Wextra fix for NamedTensor.cpp (#66897)

r-barnes committed 2 years ago

fcfa0658

[numpy] add torch.argwhere (#64257)

kshitij12345 committed 2 years ago

462f333c

Do not generate not_implemented error for forward AD when input with tangent passed to non-differentiable function (#66926)

soulitzer committed 2 years ago

892ac08a

Codegen addmm, bmm, baddbmm (#66998)

wconstab committed 2 years ago

Verified bacfd9bc

Make Lazy shape/dtype inference functions reusable (#67025)

wconstab committed 2 years ago

Verified 3b0853ad

Automated submodule update: tensorpipe (#65353)

facebook-github-bot committed 2 years ago

062ae8df

Newer Older