strisunshinewentingwang/pytorch

Pull Requests Commits

ENH Adds no_batch_dim support for maxpool and unpool for 2d and 3d (#61984)

thomasjpfan committed 3 years ago

1ec6205b

Fix type promotion for cosine_similarity() (#62054)

jbschlosser committed 3 years ago

f4ffaf0c

Improve MHA docs (#61977)

jbschlosser committed 3 years ago

e408af08

[Static Runtime] Add is_frozen to StaticModule ctor (#62020)

Hao Lu committed 3 years ago

cf3cc01f

[clang-tidy] Fix unknown GNU flag error (#62128)

1ntEgr8 committed 3 years ago

fa11103c

MAINT Migrates multilabel_margin_loss from THC to ATen (CUDA) (#60708)

thomasjpfan committed 3 years ago

9730d91a

[profiler] Nvtx support (#61634)

Ilia Cherniavskii committed 3 years ago

a6c6fd92

Smart Decay for Adam - DPER3 (#62058)

Jamie King committed 3 years ago

812bc1dd

Implement NumPy-like `frombuffer` tensor constructor. (#59077)

ysiraichi committed 3 years ago

5224490a

[Static Runtime] Fix broken test_static_runtime build (#62098)

Mike Iovine committed 3 years ago

ec4e6181

[skip ci] Refactor CIFlow init logic (#62102)

janeyx99 committed 3 years ago

b820493c

Remove redundant `torch.cuda.set_device(self.rank)` (#62097)

Yi Wang committed 3 years ago

71cfbc45

Remove duplicated movedim implementation (#61939)

peterbell10 committed 3 years ago

5ef667a8

remove `randn?` from `torch.testing` namespace (#61840)

pmeier committed 3 years ago

10ccc5a8

OpInfo Ref: fmod, remainder (#61527)

krshrimali committed 3 years ago

cb47d1f9

don't allow alias dispatch keys to go in the DispatchKeySet (#61771)

bdhirsh committed 3 years ago

c9b71549

Throw RuntimeError when numpy() is called on a tensor with conjugate or negative bit set (#61925)

anjali411 committed 3 years ago

143ef016

[special] alias for mvlgamma (#61633)

kshitij12345 committed 3 years ago

943ca5f6

[torchelastic] Improve process termination logic (#61602)

aivanou committed 3 years ago

0c55f1bd

Remove default arguments before calling to __torch_dispatch__ (#61123)

ezyang committed 3 years ago

e42360d5

Support for reference convert_fx working on gpu

HDCharles committed 3 years ago

32d0c3e8

BatchNorm: fix mixed precision usage with affine=False (#61962)

peterbell10 committed 3 years ago

0df1679e

Ignore LNK4099 for debug binary libtorch builds (#62060)

janeyx99 committed 3 years ago

e318058f

ns for fx: expose hook to define custom weight extraction functions (#62047)

vkuzo committed 3 years ago

04c95a06

ns for fx: fix typing issue in weight extraction (#62041)

vkuzo committed 3 years ago

07c6a120

ns for fx: change weight extraction to direct mapping (#62038)

vkuzo committed 3 years ago

eaba16d6

Fix some sign comparisons (#61849)

r-barnes committed 3 years ago

8a2c525d

Migrate scheduled jobs debuggability to GHA (#62056)

janeyx99 committed 3 years ago

9d405646

[DDP Comm Hook] Use a single tensor instead of a tensor list as the comm hook result (#62074)

Yi Wang committed 3 years ago

b03b45af

`clamp`: port to structured kernel (#61361)

Meghan Lele committed 3 years ago

1d2ea76a

Newer Older