strisunshinewentingwang/pytorch

Pull Requests Commits

Vectorized specialization of max_pool2d for channels-last layout (#25676)

James Reed committed 5 years ago

f7bcba33

Make tensor key in Dict works in serialization (#25442)

wanchaol committed 5 years ago

ed643382

Migrate digamma and polygamma from the TH to Aten (CUDA) (#25662)

ifedan committed 5 years ago

d939ee2d

Add CosineAnnealingWarmRestarts to optim documentation (#25421)

TortillasAlfred committed 5 years ago

38e47663

Improve handling of mixed-type tensor operations (#22273)

nairbv committed 5 years ago

88e4cee3

Enable jit fusion on ROCm (#22872)

iotamudelta committed 5 years ago

9c5a8997

add __getitem__ to class types (#25664)

Elias Ellison committed 5 years ago

82c8949a

Move most BUILD_NAMEDTENSOR macros out of header areas (#25721)

zou3519 committed 5 years ago

76bc44fb

Finish testing code examples in the docs (#25668)

davidriazati committed 5 years ago

0be29ee2

Enable two tests that were skipped b/c of rocThrust bugs fixed in ROCm 2.7

iotamudelta committed 5 years ago

c6dd4036

Cyclical learning rate multiplier: use fabs(base_lr) (#25628)

Swati Rallapalli committed 5 years ago

1559c644

Revert D17199043: [JIT] preserve ignored function return value type

suo committed 5 years ago

11eb8ac2

Align AliasInfo's operator<< with FunctionSchema (#23206)

houseroad committed 5 years ago

a294e157

Only default USE_DISTRIBUTED=True on Linux (#25725)

pietern committed 5 years ago

ce3b81fd

rocBLAS deprecated the last two parameters. (#25726)

iotamudelta committed 5 years ago

30aef56e

bring back skipped bitwise dispatch (#25689)

Ailing Zhang committed 5 years ago

bc2a37b2

Make SparseNormalize backwards compatible (#25660)

ffjiang committed 5 years ago

3be1745b

Adding RRef as return value for builtin operators (#25169)

mrshenli committed 5 years ago

197fd4f7

move USE_STATIC_DISPATCH from CI script to master cmake (#25696)

ljk53 committed 5 years ago

99b6472d

rename 'mobile' to 'static_dispatch' (#25695)

ljk53 committed 5 years ago

17e7079a

Inserting observers for all methods called in forward (#25503)

jerryzh168 committed 5 years ago

99cd83ea

Updating submodules

svcscm committed 5 years ago

7333a8c6

preserve ignored function return value type (#25262)

Elias Ellison committed 5 years ago

df043cd4

Rename FBGEMM quantized operators to generic quantized ops (#25678)

supriyar committed 5 years ago

61819260

Fix named tensor build (#25673)

zou3519 committed 5 years ago

50cb4864

Build torch.distributed with Gloo backend on macOS (#25260)

pietern committed 5 years ago

3556bea5

move GetDimFromOrderString to caffe2/core/types.h (#25671)

ljk53 committed 5 years ago

a3d0abf7

move legacy deserialization code into jit/import_legacy.cpp (#25649)

ljk53 committed 5 years ago

a35a63b8

clean up binaries/cmake for mobile (#25651)

ljk53 committed 5 years ago

3363ec92

change shape for some ops to reduce variance (#25686)

hl475 committed 5 years ago

d4226392

Newer Older