strisunshinewentingwang/pytorch

Pull Requests Commits

Exclude undefined tensors in the result of Module::parameters() / named_paramters() / buffers() / named_buffers() (#30626)

Will Feng committed 4 years ago

18ec4632

Fix typos (#30606)

bwignall committed 4 years ago

e7fe64f6

Add the explicit per-tensor/per-channel quant info when we print the module (#30591)

jianyuh committed 4 years ago

0bebfe21

Fix serialization memory lifetime issue. (#30603)

jjlilley committed 4 years ago

4dab29a2

Fix TCPStoreTest and improve tcputils::connect() (#30354)

pritamdamania committed 4 years ago

db81e13d

Disable implicit conversion warning (#30529)

WenleiHe committed 4 years ago

9e3d1941

Add support for converting quantized AvgPool2d and Reshape operations (#30490)

supriyar committed 4 years ago

968c0d4a

Add barriers to fix flaky test_graph_for_py_nested_call and (#30624)

pritamdamania committed 4 years ago

2d0a4e42

PRAGMA missing for clang (#30351)

EscapeZero committed 4 years ago

98ab55fc

Add pickler support for Device (#30131)

davidriazati committed 4 years ago

9c02b887

Add TOC to CONTRIBUTING.md (#29671)

davidriazati committed 4 years ago

19b7d49f

Turn off scalar_checks for exp, cos, cosh, tan, atan, tanh, erf, erfc. (#30434)

gchanan committed 4 years ago

56972952

Back out "Back out "Revert D18542342: Boxed variable dispatch""

smessmer committed 4 years ago

90821230

windows build (#30556)

Mingbo Wan committed 4 years ago

3636cb03

make the order btw div and mul in adagrad update consistent (#30449)

jspark1105 committed 4 years ago

d32f261f

Use pybind11::gil_scoped_* functions instead of AutoGIL/AutoNoGIL (#30274)

ezyang committed 4 years ago

1111a6b8

Update magma to 2.5.1 for Windows and switch CUDA in CI to 9.2

peterjc123 committed 4 years ago

6deb41c8

add small input shapes to some ops (#30617)

mingzhe09088 committed 4 years ago

b68d1fc3

Fix CPU_INTEL flag error on windows (#30564)

liooil committed 4 years ago

8ee61e0b

Temporarily disable test_numerical_consistency_per_tensor (#30600)

jjlilley committed 4 years ago

e6000a7c

Disable test_backward_per_tensor in test_fake_quant (#30594)

jjlilley committed 4 years ago

c780610f

Don't build test_cpp_rpc if torch is built without distributed support (#30587)

peterbell10 committed 4 years ago

53785771

Add examples to RRef doc

mrshenli committed 4 years ago

dd52f50f

Make doc source format consistent in rpc/init.cpp

mrshenli committed 4 years ago

30d70d53

Reorganize rpc API doc and add introduction (#30491)

mrshenli committed 4 years ago

ec5e4716

Improve process_group_agent() serialization speed (#29785)

jjlilley committed 4 years ago

f4e7e903

Add local shutdown to process group agent (#30330)

rohan-varma committed 4 years ago

1350b99d

Skip undefined tensors when moving torch::nn module to a different device (#30523)

Will Feng committed 4 years ago

7ac8efa6

Back out "Revert D18542342: Boxed variable dispatch"

smessmer committed 4 years ago

640109ae

Ignore logical_and and logical_or in op BC check for now (#30537)

houseroad committed 4 years ago

87f29557

Newer Older