strisunshinewentingwang/pytorch

Pull Requests Commits

[quant] Quantizable MultiheadAttention (#49866)

Zafar Takhirov committed 3 years ago

b8584b88

Remove unnecessary statement in `capture_stderr` (#52366)

Ansley Ussery committed 3 years ago

440fddf0

[Dist Profiling] Enable dist profiling for DDP (gloo only) (#52031)

rohan-varma committed 3 years ago

6dabe0b2

[PyTorch] Devirtualize TensorImpl::storage() (#51050)

swolchok committed 3 years ago

059ee85c

Fix complex acos edge cases (#52287)

malfet committed 3 years ago

4305609d

Revert D26263480: [Pytorch, Sparsity] Integrate sparse qnnpack operator in framework

malfet committed 3 years ago

72d1ccd3

[JIT] Add support for default argument values to Torchbind (#51253)

Meghan Lele committed 3 years ago

cbede834

BFloat16: enable prepacked weights's inference (#48922)

XiaobingSuper committed 3 years ago

324c6aad

[tools] Use anonymous access to access S3 bucket (#52338)

malfet committed 3 years ago

e36a900e

[PyTorch] Don't read 1 char per iteration in Unpickler::readString (#51901)

swolchok committed 3 years ago

0e2520ba

[PyTorch] Fix return value of IValue::to for Tensor/String (#51463)

swolchok committed 3 years ago

b2aa63f1

[PyTorch] Remove reference_cast in make_boxed_from_unboxed_functor (#51319)

swolchok committed 3 years ago

a9f5e722

[PyTorch] Debug-gate static_assert in KernelFunction::makeFromUnboxedFunctor (#51367)

swolchok committed 3 years ago

c442776f

Mypy fixes for pytorch master (#52090)

guilhermeleobas committed 3 years ago

975d9f25

[BE][typing] add caffe2/torch proto stubs (1 of 2) (#52341)

smacke committed 3 years ago

a8885ee7

Automated submodule update: FBGEMM (#52354)

facebook-github-bot committed 3 years ago

99619ea3

Support AST rewriting for submodules (#52297)

Ansley Ussery committed 3 years ago

d8bb9322

[Pytorch, Sparsity] Integrate sparse qnnpack operator in framework

kimishpatel committed 3 years ago

87ebaa4e

[Pytorch] Add python binding to use mobile cpu allocator. (#52323)

kimishpatel committed 3 years ago

a6e94d27

Benchmark for torch.ops.quantized.linear_prepack_fp16 operator (#52229)

Vucibatina committed 3 years ago

4501b52f

[PyTorch] Use real if constexpr behind macro in hot template (#51368)

swolchok committed 3 years ago

6e1a5b11

[PyTorch] Avoid some extra intrusive_ptr<Tuple> copies in Unpickler (#51902)

swolchok committed 3 years ago

680c4ce1

[TorchScript] C++ interface of to_<backend> (Re-land) (#52340)

malfet committed 3 years ago

f235c65a

torchvision hipify revamp fix (#51453)

rraminen committed 3 years ago

8c185e62

Automated submodule update: FBGEMM (#52255)

facebook-github-bot committed 3 years ago

35b0560e

[nnc] Add dummy reference to llvm::cfg::Update<BasicBlock> (#52321)

bertmaher committed 3 years ago

bb9e0c62

reland - ns for fx - stubs of the three APIs (compare weights, activations, activations with shadow) (#52302)

vkuzo committed 3 years ago

bfc7e281

[static runtime] use NNC to generate logit, relu and tanh (#52322)

bwasti committed 3 years ago

fa393b56

[nnc] Allow 1 ulp tolerance in log approximation (#52165)

bertmaher committed 3 years ago

41565883

Check kernel launches in caffe2/operators (#52240)

r-barnes committed 3 years ago

9409a3a3

Newer Older