strisunshinewentingwang/pytorch

Pull Requests Commits

Port `sort` to structured kernels.

ysiraichi committed 2 years ago

3e10fe32

Update Dataloader with default parameter device (#65402)

jeejakp12 committed 2 years ago

45bbc4c0

[kineto] ClientInterface stub for ProfilerKineto (#75525)

chaekit committed 2 years ago

77665e9a

[complex32] add chalf alias for complex32 and chalf method

kshitij12345 committed 2 years ago

aa51704c

[pytorch] Relax the check that makes sure number of outs equals number of returns (#76049)

larryliu0820 committed 2 years ago

128dd6b1

[NVFuser] document _jit_set_nvfuser_skip_node_kind

davidberard98 committed 2 years ago

a8ed6950

add channels last (2d) support for mkldnn_convolution (#55584)

mingfeima committed 2 years ago

92a9c0e3

Add more nvfuser merge_rules.json

davidberard98 committed 2 years ago

cebdca41

[pytorch][cupti profiler 6/n] Changes to configure Kineto cupti profiler from pytorch profiler interface (#75616)

Brian Coutinho committed 2 years ago

8385e06b

fx quant: add quantized Softmax workflow integration (#75106)

vkuzo committed 2 years ago

35545d85

SherlockNoMad committed 2 years ago

f3068dd3

sup torch script parameterlist

wukong1992 committed 2 years ago

91e9fcf5

Fix autograd.functional tests to not fail with logging tensor

soulitzer committed 2 years ago

81722f66

clarify hardtanh's definition

XiaobingSuper committed 2 years ago

27014a86

[GHF] Improve failures debugability

malfet committed 2 years ago

f31d5182

Explicitly import functional into the torch.nn namespace

willfrey committed 2 years ago

90d6e25b

remove inverse from LOBPCG

nikitaved committed 2 years ago

7f66b4ea

[SR] Fix broken unit test build (#76111)

Mike Iovine committed 2 years ago

b6a42340

[Dynamic RPC] Allow existing ranks to communicate with newly joined ranks

H-Huang committed 2 years ago

8646e0dc

[ONNX] Raise exception for unimplemented ops for non-caffe2 builds

Thiago Crepaldi committed 2 years ago

690bc1c5

Remove `-Wno-unused-variable` from `utils.cmake` (take 2) (#75538)

malfet committed 2 years ago

f6c275f5

Corrected documentation for supported padding

vitrioil committed 2 years ago

29b004be

fix VF import into torch's all field

albanD committed 2 years ago

638055df

Change default TLS behavior in dispatch to favor is-a style

albanD committed 2 years ago

cd0591df

Run Black on all of tools/

ezyang committed 2 years ago

a11c1bbd

Remove 11.5 periodic

janeyx99 committed 2 years ago

ae864d4f

back to fetch depth 0

clee2000 committed 2 years ago

5477f0ae

Modify unique to return sizes when dim is zero-length

ryandaryl committed 2 years ago

2625dab1

[DataPipe] Improving debug message when exceptions are raised within IterDataPipe

NivekT committed 2 years ago

116d0bec

[quant][core][gpu][cudnn] Added support for nhwc tensors in quantized cudnn add_relu op (#75806)

dzdang committed 2 years ago

e20793b0

Newer Older