strisunshinewentingwang/pytorch

Pull Requests Commits

Simple distributed optimizer (#29304)

aazzolini committed 4 years ago

b0cf43b2

F::embedding, F::embedding_bag, moved Embedding and EmbeddingBag options to embedding.h in options

anjali411 committed 4 years ago

604fc9ec

explicitly provide memory format when calling to clone() at ProcessGroupGloo.cpp

ifedan committed 4 years ago

65f3b98c

Properly shutdown RPC even in the case of `clean_shutdown=False`. (#29148)

pritamdamania committed 4 years ago

310343e9

move type inference for arange into c++ (#27629)

eellison committed 4 years ago

e01fc56e

Updating submodules

svcscm committed 4 years ago

9de0b635

Declare Dimname's kWildcard as extern instead of static (#29384)

peterbell10 committed 4 years ago

72eff002

Delete USE_CUDA macro use from data_parallel.h (#29483)

ezyang committed 4 years ago

344e7c26

Give a better error message when people accidentally use unsupported devices (#29409)

ezyang committed 4 years ago

b141754b

Move torch.cuda's atfork handler into C++ (#29101)

peterbell10 committed 4 years ago

bb119d95

Support softmax with D == 0 (#29167)

olittle committed 4 years ago

be757957

Fix missing space in lr_scheduler warning msg

adamjstewart committed 4 years ago

23483406

Disable flaky RPC tests (#29485)

mrshenli committed 4 years ago

3e5af226

Make smoke tests depend on s3 html update, to avoid race condition. (#29481)

ezyang committed 4 years ago

f6f428b6

Fix BC CI (#29533)

houseroad committed 4 years ago

46c4ae57

Revert D18286473: Use NNPACK for strided convolutions.

houseroad committed 4 years ago

466ab93e

Use handle pool to manage cuparse handles (#29426)

zasdfgbnm committed 4 years ago

2032482e

qavgpool benchmarking

z-a-f committed 4 years ago

5c9eae07

Adding short tests

z-a-f committed 4 years ago

958d0cd4

Use NNPACK for strided convolutions. (#29084)

Ashkan Aliabadi committed 4 years ago

5ba92097

Fix writeable strings warnings.

dlibenzi committed 4 years ago

cc6af459

nll_loss (cpu): Simplify index checking: rely on exception propagation in parallel_for (#29454)

andreaskoepf committed 4 years ago

86fee25d

Migrate `nll_loss2d` from TH to ATen (CPU) (#28304)

andreaskoepf committed 4 years ago

c7ed89cf

qpool benchmarking

z-a-f committed 4 years ago

a47fe407

Adding inplace quantized relu6

z-a-f committed 4 years ago

aa658a2a

Added all binary arithmetic tests in QFunctional

z-a-f committed 4 years ago

48741208

quantized comparators benchmarking

z-a-f committed 4 years ago

687ea746

qadd benchmarking

z-a-f committed 4 years ago

fb2eb019

set the no_deadline for the adaptive_avg_pool_nhwc test (#29502)

lly-zero-one committed 4 years ago

f5074cca

Migrate acos from TH to ATen (CUDA) (#29323)

xuhdev committed 4 years ago

6c020673

Newer Older