strisunshinewentingwang/pytorch

Pull Requests Commits

Quote in setup-ci-env (#58637)

malfet committed 3 years ago

c4f0c5ee

Fix GIL issue when acquiring multiple sessions. (#58584)

Ailing Zhang committed 3 years ago

8615fd65

Make torch::deploy work with or without cuda (#58493)

wconstab committed 3 years ago

24786bd6

port `sgn` to structured (#58197)

eqy committed 3 years ago

fbc235c2

Port fmax & fmin to structured kernel (#58458)

Kiyosora committed 3 years ago

b5e39bce

[FX Splitter] dump final graph and print operator stats via to_glow API

Luoshang Pan committed 3 years ago

e179a568

refactor ASGD to use functional API (#58410)

iramazanli committed 3 years ago

9a622f4c

remove redundant getDispatchKeySetUnboxed(eligibleKeys) (#58535)

Basil Hosmer committed 3 years ago

208b36f1

Rename namespace `vec256` to `vec`, struct `Vec256` to `Vectorized` (and other related classes/structs) (#58438)

imaginary-person committed 3 years ago

47c566eb

Revert D28461013: [nnc] Enable CPU fusion inside Facebook, take 2

bertmaher committed 3 years ago

a6b358d5

[FX] Add APIs to mutate specific args/kwargs (#58571)

James Reed committed 3 years ago

36adc3f0

[THC] Rename THCTensorMathMagma from cu to cpp (#58521)

malfet committed 3 years ago

296d2a43

Added publishing of test results and minor fixes to Az DevOps Build Logic (#58436)

mstfbl committed 3 years ago

ae99640a

[profile] fix recorded data type (#58531)

kingchc committed 3 years ago

b9b8522e

Revert "Move Azure MultiGPU tests back to nightly (#58242)" (#58451)

skyline75489 committed 3 years ago

8de8b492

Fix some tensor operators to return `NotImplemented` for invalid inputs (#58216)

asi1024 committed 3 years ago

3113a1de

step 0 of cuDNN v8 convolution API integration (#51390)

zasdfgbnm committed 3 years ago

6c70cbed

[ATen][Quant] Pass at::Tensor by reference (#58284)

Hao Lu committed 3 years ago

954d39ba

model_dump: Accept variable-length debug info (#57660)

dreiss committed 3 years ago

a9137543

model_dump: Use DumpUnpickler.load instead of .dump (#57659)

dreiss committed 3 years ago

ab1fdbef

model_dump: Add a section that summarizes tensor memory usage (#57658)

dreiss committed 3 years ago

53078924

model_dump: Handle dict rendering (#57657)

dreiss committed 3 years ago

ef4e6036

model_dump: Handle torch.device objects (#57656)

dreiss committed 3 years ago

72ff3163

model_dump: Refactor renderTensor into a helper method (#57655)

dreiss committed 3 years ago

a380575f

model_dump: Implement "Hider" properly (#57654)

dreiss committed 3 years ago

3ff76af2

move code to Blas.cpp, clean up THC magma (#58526)

ngimel committed 3 years ago

3f0b0816

[JIT] improve documentation (#57991)

garymm committed 3 years ago

703cfdc9

s/foward/forward/g (#58497)

Chillee committed 3 years ago

79a258f4

Added OperatorMap for mapping Operator to any template <T> (#58060)

Edvard Ghazaryan committed 3 years ago

ccad77aa

[Reducer] Remove some unused variables (#58524)

rohan-varma committed 3 years ago

1ba05efd

Newer Older