strisunshinewentingwang/pytorch

Pull Requests Commits

Revert D29883676: Migrate thnn_conv_depthwise2d from THC to ATen

ejguan committed 3 years ago

acaac70f

[2/N] Remove unittest.skip across all of torch.distributed. (#61887)

pritamdamania committed 3 years ago

82d81455

fix typo errors in quantization-support.rst Line320 (#44447)

huqinghao committed 3 years ago

7fc96db4

Reenable AMP on XLA (#61861)

ezyang committed 3 years ago

5f7f08f4

Fixing the case when starter nodes depend on get_attr node (#62234)

khabinov committed 3 years ago

a0c1c7e5

Revert D29810657: [bc-breaking] reference option for linear produce a pattern instead of reference linear module

ejguan committed 3 years ago

8cdf16d1

det_backward: correct, more robust and with complex support [clone] (#61905)

nikitaved committed 3 years ago

d7ddae8e

Migrate thnn_conv_depthwise2d from THC to ATen (#62006)

peterbell10 committed 3 years ago

de3a4eb5

[bc-breaking] reference option for linear produce a pattern instead of reference linear module (#61892)

jerryzh168 committed 3 years ago

9df60513

[7/N] Nnapi backend delegation preprocess: compile_spec sanity check (#62213)

Amyh11325 committed 3 years ago

6c6a9c73

[DDP] Log if graph is static at end of training (#61871)

rohan-varma committed 3 years ago

2cbc0ede

[Static Runtime] Enforce proper output dtype for many ops (re-land) (#62267)

Mike Iovine committed 3 years ago

79eb8bb2

Disable ccache for nccl builds (#62208)

nairbv committed 3 years ago

2eef1f27

Forward fix mypy (#62263)

ejguan committed 3 years ago

dc55d511

Add forward mode differentiation for inverse and solve (#62160)

IvanYashchuk committed 3 years ago

3cd12448

Initial ModuleInfo implementation (#61935)

jbschlosser committed 3 years ago

a0309f89

Remove faulty process group code (#61907)

H-Huang committed 3 years ago

afe36443

Revert D29887367: [Static Runtime] Enforce proper output dtype for many ops

ejguan committed 3 years ago

a3be2ecc

Create linalg and parametrizations codeowners (#62086)

lezcano committed 3 years ago

b599c1e7

[AutoAccept][Codemod][FBSourceClangFormatLinter] Daily `arc lint --take CLANGFORMAT`

CodemodService FBSourceClangFormatLinterBot committed 3 years ago

228b50e0

[bc-breaking] Produce quantization pattern for add_scalar and mul_scalar (#61859)

jerryzh168 committed 3 years ago

2d7c1e3f

Add device and key for lazy tensors (#61621)

asuhan committed 3 years ago

b176feec

Add option to skip GH validation for torch.hub (#62139)

malfet committed 3 years ago

2945a73d

[DDP/Functional Optim] Support kwarg arguments (#62079)

rohan-varma committed 3 years ago

64283fe1

[Functional Optim] Test kwargs parity for SGD (#62078)

rohan-varma committed 3 years ago

c0ebeca1

Revert D29801652: Refactor Tensor::to to call a primitive that is not copy_.

malfet committed 3 years ago

478098aa

Parity tests for functional optimizer step_param (#61756)

rohan-varma committed 3 years ago

69adb219

Fix infinite loop in `_validate_not_a_forked_repo()` (#62072)

malfet committed 3 years ago

b6d10a3a

[PyTorch][Edge] Serializing sub modules with same names (#61933)

pavithranrao committed 3 years ago

d0f43092

DOC: remove git stamp from release documentation version (#58486)

mattip committed 3 years ago

a13f714b

Newer Older