strisunshinewentingwang/pytorch

Pull Requests Commits

Fix a CUDA memory leak in MultiLabelMarginCriterion error checking. (#30767)

gchanan committed 4 years ago

473a0448

Turn off scalar_check for is_target for MultiLabelMarginCriterion, which is handled correctly in code. (#30766)

gchanan committed 4 years ago

ba1a9871

Support 0-d tensors in CUDA MultiLabelMarginCriterion. (#30765)

gchanan committed 4 years ago

35a69978

Provide full path for buck hipification (#30746)

Serhat Yilmaz committed 4 years ago

c4e9748b

CUDA-strided-complex Binary and Unary Op support (#30295)

dylanbespalko committed 4 years ago

f2a2fec4

Clean up non-C++14 code (#28443)

smessmer committed 4 years ago

139aa519

fix AvgPool2d for 2^31-1 sized inputs, and get test_cuda_kernel_loop_… (#30771)

soumith committed 4 years ago

a939b52d

Make `InsertQuantDeQuantHelper` global (#30550)

jerryzh168 committed 4 years ago

1d20c32b

Supporting making submodules unique (#30037)

jerryzh168 committed 4 years ago

c4c2e233

Stop producing op_version_set version numbers.

zdevito committed 4 years ago

7a2889b0

Invoke more passes in `insertObservers` (#30473)

jerryzh168 committed 4 years ago

3c1bb21c

Back out "make the order btw div and mul in adagrad update consistent" (#30737)

jspark1105 committed 4 years ago

e09c4153

Don't install pybind11 header directory for system pybind11 installs (#30758)

ngoldbaum committed 4 years ago

1f1ce53e

fix anynonzero op

wanchaol committed 4 years ago

569ea63f

Updating submodules

svcscm committed 4 years ago

1d8a1314

Updating submodules

svcscm committed 4 years ago

cd032c7f

AddConstant and findConstant for ClassType (#29217)

jerryzh168 committed 4 years ago

17077744

Improve documentation around builtin functions (#30347)

davidriazati committed 4 years ago

2308a0ec

Kill THNN version of MultiMarginCriterion; it's not used anymore.

gchanan committed 4 years ago

42e79d7e

Add the __torch_function__ API override mechanism (#30730)

ngoldbaum committed 4 years ago

9d3402e4

Move Tanh backward to Aten(CPU+CUDA) (#30224)

XiaobingSuper committed 4 years ago

289e9a07

Cache compilation of free functions (#30503)

Elias Ellison committed 4 years ago

d38f9117

add MaskedRowWiseSparseAdagrad

jspark1105 committed 4 years ago

9d69c55b

Move scalar_check logic from codegen to code in NLLLoss. (#30670)

gchanan committed 4 years ago

786de338

Simplify scalar_check of nll_loss. (#30669)

gchanan committed 4 years ago

fa2aa245

Move scalar_check for total_weight in NLLLoss functions to code from codegen. (#30665)

gchanan committed 4 years ago

6918f0ce

Rename QuantizeHelper to InsertQuantDeQuantHelper (#30549)

jerryzh168 committed 4 years ago

756f279d

InsertObservers for shared class types (#30548)

jerryzh168 committed 4 years ago

f73cd280

add irregular c10 op registration/invocation cases to test project (#30558)

ljk53 committed 4 years ago

6e145b46

Check the error return of nvrtcGetProgramLogSize and nvrtcGetProgramLog (#30663)

ezyang committed 4 years ago

a55f125e

Newer Older