strisunshinewentingwang/pytorch

Pull Requests Commits

Set USE_KINETO=1

ilia-cher committed 3 years ago

ac2cdb6d

[quant] Add non-fbgemm fallback implementation for embedding lookup ops (#50706)

supriyar committed 3 years ago

e0096650

[FX] Fix NoneType annotation in generated code (#50777)

James Reed committed 3 years ago

5205cc1c

[JIT] Print out CU address in `ClassType::repr_str()` (#50194)

Meghan Lele committed 3 years ago

8f5ad00e

Cat benchmark: use mobile feed tensor shapes and torch.cat out-variant (#50778)

Marat Subkhankulov committed 3 years ago

dea9af5c

Generalize `sum_intlist` and `prod_intlist`, clean up dimensionality functions (#50495)

r-barnes committed 3 years ago

06c734d8

Fix Native signature for optional Tensor arguments (#50767)

smessmer committed 3 years ago

47c57b88

Fix USE_MKLDN defaults (#50782)

malfet committed 3 years ago

cebab83d

Add Sparse support for torch.sqrt (#50088)

hlahkar committed 3 years ago

4ff1823f

[FX] Fix tracing a free function with embedded constant (#50639)

James Reed committed 3 years ago

38c45bdd

Revert "Stable sort for CPU (#50052)" (#50752)

Xinyu Li committed 3 years ago

7526e38c

Automated submodule update: tensorpipe (#50765)

facebook-github-bot committed 3 years ago

08c90d9e

Fix bug in hipify if include_dirs is not specified in setup.py (#50703)

jithunnair-amd committed 3 years ago

327539ca

whitelist ops we can build shapes for (#49125)

Krovatkin committed 3 years ago

526659db

Fix nvcc function signature causing assert in TypeIndex.h (#49778)

zfenice committed 3 years ago

4816bf62

add type annotations to torch.nn.modules.container (#48969)

guilhermeleobas committed 3 years ago

a9e46f14

Better split of the windows test jobs (#50660)

peterjc123 committed 3 years ago

a1b1d0cd

initial commit to enable fast_nvcc (#49773)

Rong Rong (AI Infra) committed 3 years ago

ebd142e9

Remove instance of blacklist (#50478)

r-barnes committed 3 years ago

f7b2b22b

Disable tracer warning for slicing indices. (#50414)

tugsbayasgalan committed 3 years ago

0c9fb4af

[FX] Fix using fx.wrap as a decorator (#50677)

jansel committed 3 years ago

3344f061

Remove workaround for TensorPipe failing to get device of CUDA ptr (#50580)

lw committed 3 years ago

05036564

Fix caffe2 import tools.codegen (#50353)

KumaTea committed 3 years ago

5f33f223

Shouldn't need user local install for ROCm build (#50299)

jithunnair-amd committed 3 years ago

a9deaf36

Update TensorPipe submodule (#50733)

lw committed 3 years ago

cad47531

Clean up complex autograd test list (#50615)

anjali411 committed 3 years ago

4511f2cc

Consolidate mypy tests and args (#50631)

samestep committed 3 years ago

937eff58

Striding for lists Part 1 (#48719)

tugsbayasgalan committed 3 years ago

1a38fa99

Add instructional error message for cudnn RNN double backward workaround (#33884)

zou3519 committed 3 years ago

1154a859

Add complex support for `torch.{acosh, asinh, atanh}` (#50387)

anjali411 committed 3 years ago

5d64658c

Older