strisunshinewentingwang/pytorch

Pull Requests Commits

Fix DistributedDataParallelTest.test_accumulate_gradients (#20351)

pietern committed 5 years ago

a0e5240a

Remove const_cast's from subgraph matcher. (#20303)

Owen Anderson committed 5 years ago

02df1ccd

Adding setup job as prereq to html update jobs

pjh5 committed 5 years ago

e47b2100

Remove SourceLocation (respin) (#20333)

zdevito committed 5 years ago

3afd9968

Make DistributedDataParallel usable with CPU models (#20236)

pietern committed 5 years ago

558c6c4d

Refactor core DistributedDataParallel tests (#20235)

pietern committed 5 years ago

f32c9bd5

Add c10d::broadcast_coalesced and tests (#20234)

pietern committed 5 years ago

caa0d0c5

Fix crash issue in conv+sum fusion for MKLDNN on caffe2 (#20139)

gujinghui committed 5 years ago

c31fccd6

Move librosa and psutil installation from CI script to docker images build script (#20299)

bddppq committed 5 years ago

8c3a7bb5

Revert D15275731: Remote SourceLocation

wanchaol committed 5 years ago

e870b11a

Remote SourceLocation (#20300)

zdevito committed 5 years ago

eca91de5

profiler: improve repr for averaged events (#20281)

salexspb committed 5 years ago

726661b1

Re-enable CUDA tests for C++ API (#20238)

Will Feng committed 5 years ago

39af9563

Extend testAvailableArgTypes (#20185)

smessmer committed 5 years ago

fe714862

Allow Dict type in c10 operators (#20184)

smessmer committed 5 years ago

e0166f46

smessmer committed 5 years ago

c9212903

Add unit test to ensure no gradients sync when calling ddp.module(input) (#20282)

chenyangyu1988 committed 5 years ago

2019f6cd

fix typo in adaptive methods annotation (#20306)

Ailing Zhang committed 5 years ago

899bddee

Add test/test_namedtensor.py (#20168)

zou3519 committed 5 years ago

83a80d2b

Add namedtensor build and test to the CI (#20163)

zou3519 committed 5 years ago

199fa12d

Add USE_NAMEDTENSOR compilation flag. (#20162)

zou3519 committed 5 years ago

e01a5bf2

Fix in file position logic: file descriptor and Python-side handle (#20270)

philippslang committed 5 years ago

f23fb66e

error instead of crashing on attempt to subclass typed tensors (#20283)

nairbv committed 5 years ago

c406bf20

Switch off USE_DISTRIBUTED on default for MSVC (#20302)

peterjc123 committed 5 years ago

1e35ef07

Fixed Softmax doc to specify dimension to prevent warning in 1.1.0. (#20310)

ZaydH committed 5 years ago

2aea5b63

Use torch::get/set_num_threads without additional includes beyond torch/torch.h (#20176)

Will Feng committed 5 years ago

0087069d

Sort User Defined Classes (#19706)

eellison committed 5 years ago

916ef768

Bump op_version_set (#19812)

zdevito committed 5 years ago

1102f4c5

Refactor ChunkDataReader API + fix missing headers (#19485)

Thiago Crepaldi committed 5 years ago

3d4d7b90

Eliminate some const_cast's.

resistor committed 5 years ago

bed1d7d3

Newer Older