strisunshinewentingwang/pytorch

Pull Requests Commits

Options based registration API (#20514)

smessmer committed 5 years ago

cb6be424

Add qint8 type (int8_t) (#19984)

jerryzh168 committed 5 years ago

85fad059

Add a pybind for Module::get_functions. (#20594)

Mikhail Zolotukhin committed 5 years ago

986c9eb5

Modify the Insert quant-dequant test cases to look for q-dq pattern (#20672)

nishantpdce committed 5 years ago

2af5911a

Some small performance fixes for c10 dispatcher (#20472)

Roy Li committed 5 years ago

e42665cf

Improve CPUAllocator OOM message (#20618)

Dmytro Dzhulgakov committed 5 years ago

d9dcfacd

Fix missing env for update_binary_size job

pjh5 committed 5 years ago

e79610c0

Misc error message improvements (#19369)

davidriazati committed 5 years ago

c267d0c8

SubgraphMatcher: add attributes support. (#20602)

Mikhail Zolotukhin committed 5 years ago

3be2f7c8

Use python type string for user facing error msgs (#20657)

Elias Ellison committed 5 years ago

0b9b929d

Add list(string) (#20617)

davidriazati committed 5 years ago

c819d767

Add `_enable_recursive_script` to try to script all Python functions (#19578)

davidriazati committed 5 years ago

a543586b

Add support for __getstate__/__setstate__ on module (#20242)

davidriazati committed 5 years ago

cd28ff53

Split out gpu/cpu targets based on gpu_library_targets (#20633)

lertoz committed 5 years ago

5f14ef8c

Remove unnecessary format literals from error message.

ezyang committed 5 years ago

79c5dc31

Remove TH/THC link for single matrix inverse (#20534)

vishwakftw committed 5 years ago

26dfeffa

Revert D15393514: [pytorch][PR] Refine CosineAnnealingWarmRestarts doc for issue #20028

ezyang committed 5 years ago

839a69f5

Add matmul optimization for the case A.ndim <= 2 && B.ndim >= 3 (#20448)

skrah committed 5 years ago

8c9f4c56

ir.cpp, module.cpp: clang-format. (#20592)

Mikhail Zolotukhin committed 5 years ago

a212a5b9

Don't split 256-bit AVX2 load/store intrinsics (#20609)

colesbury committed 5 years ago

b90790ab

fix WAR race (#20182)

Natalia Gimelshein committed 5 years ago

000d73cc

Refine CosineAnnealingWarmRestarts doc for issue #20028 (#20267)

Kirayue committed 5 years ago

3c69c9a7

Update documentation for CTCLoss (#20422)

zhiqwang committed 5 years ago

cfb87c10

Export sign onnx operator (#20470)

daquexian committed 5 years ago

35e0015c

Make C10_NODISCARD macro more portable for nvcc+clang. (#20324)

ezyang committed 5 years ago

4e551a7e

Remove checks for CUDA 8 in LU-based tests (#20482)

vishwakftw committed 5 years ago

690efa52

Make check-doxygen.sh output more interpretable. (#20362)

ezyang committed 5 years ago

110ed511

Enable simd and loop vectorizer with MSVC

peterjc123 committed 5 years ago

1136ad59

Emphasize all DDP forward() outputs must participate in computing loss (#20586)

mrshenli committed 5 years ago

fa4ca4e7

Fix upsample kernel launch / reorder arguments (#20505)

xmnlab committed 5 years ago

c941abbc

Newer Older