strisunshinewentingwang/pytorch

Pull Requests Commits

Clean up some imports in cuda kernel code. (#47400)

gchanan committed 3 years ago

ff3e1de6

Fix collect_env when pytorch is not installed (#47398)

malfet committed 3 years ago

848901f2

Split up BinaryMiscOpKernels.cu because it's slow to compile. (#47362)

gchanan committed 3 years ago

da491d75

[NNC] Fix an issue in Cuda fusion with fp16 scalar vars coerced to float (#47229)

nickgg committed 3 years ago

33cf7fdd

Add tests for DDP control flow models. (#47206)

rohan-varma committed 3 years ago

31c9d2ef

Add `input` argument to `autograd.backward()` cpp api (#47214)

soulitzer committed 3 years ago

2e5bfa98

Skip iomp5 emebedding if torch_cpu could not be found (#47390)

malfet committed 3 years ago

6f602518

[Pytorch] Add basic c10::optional tests (#47014)

swolchok committed 3 years ago

ae706378

[pytorch] Remove c10::nullopt_t::init (#47013)

swolchok committed 3 years ago

17be8ae1

[JIT] add freeze to docs (#47120)

Elias Ellison committed 3 years ago

7ab843e7

Expand GRADIENT_IMPLEMENTED_FOR_COMPLEX to allow named tensors (#47289)

jonasteuwen committed 3 years ago

a11bc049

Add vulkan reshape op (hack) (#47252)

SS-JIA committed 3 years ago

5d82311f

[Gradient Compression] Export sizes, along with length and offset of each variable to GradBucket for PowerSGD (#47203)

Yi Wang committed 3 years ago

6b3802a7

Renamed a TensorListMetaData property. Cleaned up a test (#46662)

Iurii Zdebskyi committed 3 years ago

2c554266

[quant][graphmode][fx] Test to make sure dequantize node are placed properly (#47332)

jerryzh168 committed 3 years ago

f588ad6a

Revert D24481801: Optimize backward for torch.repeat

ejguan committed 3 years ago

bba5a311

Fix onnx test-reports path in CI (#47315)

janeyx99 committed 3 years ago

4189c3ca

Including generator param in randperm documentation (#47231)

janeyx99 committed 3 years ago

01da0fe5

Revert "Revert D24335982: explicitly error out in comparison ops when the types don't match" (#47288)

bdhirsh committed 3 years ago

fe17269e

randperm: add torch check to ensure generator device = tensor device (#47022)

janeyx99 committed 3 years ago

e4bc785d

Removing caffe2 and third_party from our code coverage (#47310)

janeyx99 committed 3 years ago

07e8f48e

Implement copysign (#46396)

ejguan committed 3 years ago

f1ac63d3

[pt][static_runtime] Memory model (#46896)

Hao Lu committed 3 years ago

996f444c

Move python-independent c10d implementations to torch/lib (#47309)

gmagogsfm committed 3 years ago

5c4bd9a3

Support int32 indices and offsets in nn.EmbeddingBag (#46758)

Qi Zhou committed 3 years ago

0ec717c8

Expose SparseLengthsSum8BitRowwiseSparse to C10 (#47306)

houseroad committed 3 years ago

a2f9c7d4

[quant][graphmode][fx] Add support for qat convbn{relu}1d (#47248)

jerryzh168 committed 3 years ago

0cba3e37

Do not delete rpath from torch.dylib on Darwin (#47337)

malfet committed 3 years ago

3a002457

[quant][eagermode] Avoid inserting fakequant for sigmoid/hardsigmoid/tanh in eval mode (#47297)

jerryzh168 committed 3 years ago

53a5f08e

[quant][graphmode][fx][fix] Fix error that DefaultQuantizer is not inserted after a module configured with None qconfig (#47316)

jerryzh168 committed 3 years ago

c6fe65bf

Newer Older