strisunshinewentingwang/pytorch

Pull Requests Commits

Bumps readable file format version for torch.full inferring float from int values (#40089)

Mike Ruberry committed 4 years ago

3d8de74e

Revert D22071278: [quant][graphmode] Refactor dynamic quant tests

supriyar committed 4 years ago

b5d54db6

Revert D22071277: [quant][graphmode] Test JIT tracing for dynamic quant cases

supriyar committed 4 years ago

cb1a1942

Remove unecessary copy within blob serialization (#40096)

Sean Lynch committed 4 years ago

64689c24

Split TensorIteratorConfig out of TensorIterator (#39803)

ezyang committed 4 years ago

ba98c0e3

LayerNorm use Fused Multiply and Add (#40012)

venkatacrc committed 4 years ago

54c0ee1e

Fix KeypointRCNN test (#39589)

KsenijaS committed 4 years ago

da8cd826

Back out "Revert D21986243: TORCH_FN" (#40110)

smessmer committed 4 years ago

f69b72c7

[quant] Support general op modules with inplace options (#39919)

jerryzh168 committed 4 years ago

41fa4bef

[quant][graphmode] Test JIT tracing for dynamic quant cases (#40040)

supriyar committed 4 years ago

fa4244d7

[quant][graphmode] Refactor dynamic quant tests (#40039)

supriyar committed 4 years ago

ddeaa743

[quant][graphmode] Support quantizing `repeat` (#39925)

jerryzh168 committed 4 years ago

461aa8a1

Fix destructor ordering for cuda handle pools (#39345)

peterbell10 committed 4 years ago

ee365c58

Avoid using default process group in ProcessGroupAgent. (#39909)

pritamdamania committed 4 years ago

145df306

fix more duplicated names (#40062)

linbinyu committed 4 years ago

7021635d

Upgrade DNNL to 1.5 (#40088)

pinzhenx committed 4 years ago

7f270233

Revert D22069566: Revert D22013026: [quant][graphmode] Pass debug option into insert_quant_dequant pass

jerryzh168 committed 4 years ago

ec1833bc

Remove global CMAKE_INSTALL_RPATH_USE_LINK_PATH directive (#37737)

mattip committed 4 years ago

49732f04

Remove item and data_ptr for std::complex (#39838)

zasdfgbnm committed 4 years ago

d57ca73c

[quant][graphmode] Support squeeze/unsqueeze (#39924)

jerryzh168 committed 4 years ago

181ea1ac

[xplat] Add Windows specific ATen build definitions (#40092)

malfet committed 4 years ago

f1a5f661

[JIT] remove callable check to compile objects with __call__ (#40041)

Elias Ellison committed 4 years ago

b3dd4d9c

Revert D20496044: [pytorch][PR] Change AccumulateGrad to yield `.grad`s that match weights' memory layout

albanD committed 4 years ago

f1e575a0

optimize upsample performance linear mode on CPU (#34864)

mingfeima committed 4 years ago

4b5530de

[TensorPipe] Fix transport/channel priorities (#40090)

lw committed 4 years ago

5843854e

DOC: fix rpc reference in top-level index (#40077)

mattip committed 4 years ago

dd581b45

Batching rule for torch.mul (#39859)

zou3519 committed 4 years ago

56b4b441

[JIT] Add registry for backend lowering functions (#39552)

Meghan Lele committed 4 years ago

33b82c72

Reduce memory requirement for test_argminmax_large_axis (#40036)

peterbell10 committed 4 years ago

ad86c94f

Revert D21628596: Refactor LSTM tests

raghuramank100 committed 4 years ago

5add2e86

Newer Older