strisunshinewentingwang/pytorch

Pull Requests Commits

Revert D13551909: [fbcode] logdevice for generic feature type

bddppq committed 5 years ago

52135e9b

logdevice for generic feature type (#16191)

huginhuangfb committed 5 years ago

11a2b379

layernorm example (#16350)

bwasti committed 5 years ago

265ed8ff

plug caffe2 into jit (#16331)

bwasti committed 5 years ago

6d2aee4a

Add RunOperator for using FunctionSchema registered ops easily in caffe2 (#16173)

bwasti committed 5 years ago

d4b60f40

Add correct Input() shim to caffe2 operator impl (#16048)

bwasti committed 5 years ago

3b6b777a

Relax lower bound for nogil timing test to avoid false alarm (#16259)

mrshenli committed 5 years ago

7ce634eb

Code-style fixes. (#16342)

Mikhail Zolotukhin committed 5 years ago

c787de62

disable testing group conv with EIGEN engine (#16335)

jspark1105 committed 5 years ago

6700eff0

Remove unneeded manual unwrap optionals (#16245)

Elias Ellison committed 5 years ago

c2be9f14

fix buildindexop (#16341)

sunnieshang committed 5 years ago

f769cf99

Revert D13747581: Optimize SpatialBN on GPU

dinhvh committed 5 years ago

69b5ae4c

Add Test for ReinitializeTensor (#16338)

jerryzh168 committed 5 years ago

0b470d0a

Add thread-local guard: at::AutoNonVariableTypeMode (#15939)

Will Feng committed 5 years ago

2a70f24c

reduce parameter space of test_1x1_conv to avoid timeout (#16223)

jspark1105 committed 5 years ago

f0dd85d1

Update docs to include variable annotation example (#16324)

Sidney Zhang committed 5 years ago

fdda533e

Delete duplicate copy of THCCachingAllocator. (#16226)

ezyang committed 5 years ago

792cb774

Move THCCachingAllocator to c10_cuda. (#16119)

ezyang committed 5 years ago

e936a690

Remove unnecessary includes and headers from THCCachingAllocator, move to at::cuda:: namespace (#16117)

ezyang committed 5 years ago

24b50f14

Directly include headers from ATen.

Mikhail Zolotukhin committed 5 years ago

47bf3066

Refactor the docs build workflow (#16265)

zou3519 committed 5 years ago

af513cd4

Save a little bit of work in constant pooling by not moving nodes that will get deleted.

Owen Anderson committed 5 years ago

a4be1537

Handle non-contiguous inputs with mkldnn convolution. (#16300)

gchanan committed 5 years ago

0cb24098

Optimize SpatialBN on GPU (#16202)

xiaomengy committed 5 years ago

45c3cc91

optimize group_norm (#16216)

xiaomengy committed 5 years ago

60241e94

Fix the tensor deserialization problem of jit script module on CUDA (#16279)

houseroad committed 5 years ago

8ab4d348

Small fixes for pdist (#16210)

Erik Brinkman committed 5 years ago

3cba115a

Fix comparison in ReinitializeTensor (#16294)

jerryzh168 committed 5 years ago

0a3932ac

Fix issues under caffe round 1

Benny Chen committed 5 years ago

f25322fb

Add support for overloaded functions (#15556)

David Riazati committed 5 years ago

31de19f2

Newer Older