strisunshinewentingwang/pytorch

Pull Requests Commits

[PyTorch] Remove StringView from RecordFunction interface (1/2) (#68410)

swolchok committed 2 years ago

1d84d8c5

Use `cub::FutureValue` to simplify 64bit indexing split of cub scan (#66711)

zasdfgbnm committed 2 years ago

22690c2c

[vulkan] Update VMA settings to reduce memory usage (#69088)

SS-JIA committed 2 years ago

c48e6f01

[DDP][BE][Docs] Clarify checkpoint support (#68827)

rohan-varma committed 2 years ago

fcd1375b

Refactor DDP checkpoint tests (#68792)

rohan-varma committed 2 years ago

994f110a

[JIT] internal build bug fix (#69061)

makslevental committed 2 years ago

49abda20

[quant][embedding qat] Set FakeQuant zeropoint dtype matches observer (#68390)

b-koopman committed 2 years ago

5e0302e1

ammend tensors.rst and torch.rst for doc generation (#69030)

nikulpatel3141 committed 2 years ago

8f9f5594

[fx] remove CPatcher (#69032)

suo committed 2 years ago

0aa9d177

Markdown was linking to repo rather than pytorch.org website (#68937)

iwishiwasaneagle committed 2 years ago

81246ed0

Revert D32706197: Sparse: Implement simple unary ufuncs operators

seemethere committed 2 years ago

251686fc

Remove finput from slow2d signatures (#68896)

jbschlosser committed 2 years ago

8fef7c09

[Static Runtime] [Code Cleanup] Reduce indentation depth in ops.cpp (#69028)

d1jang committed 2 years ago

cd3e37cb

[JIT] Separate GPU implementation of frozen_conv_add_relu_fusion.cpp (#68149)

davidberard98 committed 2 years ago

cfc75c21

[static runtime] dequantize out variant (#68664)

ajyu committed 2 years ago

7342b654

Revert D32099294: Split cuda: list cpp files that go in _cu library explicitly

malfet committed 2 years ago

d3de3546

CompositeImplicitAutograd compliance testing (#65819)

zou3519 committed 2 years ago

6fea7499

[LT] Sync LTC branch changes on torch/csrc/lazy/core (#69012)

desertfire committed 2 years ago

b83e8d56

[Static Runtime] Fix fb::expand_dims schema (#68636)

Hao Lu committed 2 years ago

39ab4171

dbr quant overhead [14/x]: cache whether an op is a module (#68877)

vkuzo committed 2 years ago

5b37ac54

Split cuda: list cpp files that go in _cu library explicitly (#67216)

Michael Dagitses committed 2 years ago

b47ae981

Remove native_functions.yaml dependency from IndexKernel.{cpp,cu} (#66914)

peterbell10 committed 2 years ago

174eea8a

Remove native_functions.yaml dependency from TensorModeKernel.cu (#66913)

peterbell10 committed 2 years ago

f7d59894

[CUDA Pinned Memory] Alternative implementation of pinned memory allocator focusing on multi-threaded scalability (#68906)

Andrew Tulloch committed 2 years ago

ec1339a4

[LTC] Upstream some utilities (#69046)

Jiewen Tan committed 2 years ago

0cdeb586

Sparse: Implement simple unary ufuncs operators (#68887)

peterbell10 committed 2 years ago

fbaa19a6

[TensorExpr] Supress TracerWarnings in test_unsupported in test_jit_fuser_te.py. (#68757)

Mikhail Zolotukhin committed 2 years ago

3186d369

[TensorExpr] Allow for 'keepdim' argument in aten::mean in NNC's external call. (#68756)

Mikhail Zolotukhin committed 2 years ago

75ce0406

[TensorExpr] IRPrinter: print sizes and name when visiting a Buf. (#68755)

Mikhail Zolotukhin committed 2 years ago

a93f505e

Add option to get input dtype from user (#68751)

priyaramani committed 2 years ago

8cc9ec2f

Newer Older