strisunshinewentingwang/pytorch

Pull Requests Commits

Add Mul op for Vulkan (#47021)

SS-JIA committed 3 years ago

38265acf

Update pybind to 2.6.0 (#46415)

malfet committed 3 years ago

2b6a720e

Fix segfault with torch.orgqr. (#46700)

v0dro committed 3 years ago

2249a293

Added torch.linalg.tensorsolve (#46142)

IvanYashchuk committed 3 years ago

f629fbe2

Fix implicit conversion (#46833)

r-barnes committed 3 years ago

13b4127c

Fix DDP documentation (#46861)

rohan-varma committed 3 years ago

ecdbea77

Show old kernel location when there are mismatches (#46850)

smessmer committed 3 years ago

262bd643

Disable softmax tests on ROCm (#46793)

ashish committed 3 years ago

dfdc1dbe

Implement LengthsToOffsets operator in Caffe2 (#46590)

Brandon Lin committed 3 years ago

4a581ba6

[RFC][LocalSession] Fix workspace type

kunalb committed 3 years ago

18d273dc

[FX] Put inf and nan in globals instead of with an import string (#47035)

James Reed committed 3 years ago

d0df29ac

[RPC Framework] Support remote device format "<workername>/<device>" (#46773)

Yi Wang committed 3 years ago

cab32d9c

Throw an exception in the constructor of torchscript serialization to avoid double-exception (#44266)

iseeyuan committed 3 years ago

b553c06a

[RFC] Add OperatorHandle overload to the RecordFunction::before() method (#46401)

dhruvbird committed 3 years ago

9c1a41b7

Fix negative column numbers for the torch.eye (#46841)

RockingJavaBean committed 3 years ago

604e1b30

[numpy] `torch.cos`, `torch.tan` : promote integer inputs to float (#46706)

kshitij12345 committed 3 years ago

5c8aad11

Use f-strings in torch.utils.cpp_extension (#47025)

malfet committed 3 years ago

42a51148

[pytorch] get rid of cpp_type_str from pybind codegen (#46977)

ljk53 committed 3 years ago

9d23fd5c

[pytorch] simplify tensor options logic in pybinding codegen (#46976)

ljk53 committed 3 years ago

79474a19

add support for different memory sizes on size_based_partition (#46919)

scottxu0730 committed 3 years ago

a86b3438

[quant][graphmode][fx] Remove inplace option for convert_fx (#46955)

jerryzh168 committed 3 years ago

c2a39513

Disable test_joing_running_workers for TSAN. (#46966)

pritamdamania committed 3 years ago

ad260ae7

Fix signed-to-unsigned conversion warning (#46834)

r-barnes committed 3 years ago

9fefb406

Fix object-based collectives API to use torch.cuda.current_device instead of (#46897)

rohan-varma committed 3 years ago

c7183c98

Various cleanups to ir_emitter and friends (#46686)

suo committed 3 years ago

dc817635

[quant] Fix flaky test test_histogram_observer_against_reference (#46957)

supriyar committed 3 years ago

fc2bd991

[doc] Fix info on the shape of pivots in `torch.lu` + more info on what and how they encode permutations. (#46844)

nikitaved committed 3 years ago

cd26d027

Fix torch.version.debug generation (#47006)

malfet committed 3 years ago

058f43fc

Add Vulkan op Add. (#44017)

Ashkan Aliabadi committed 3 years ago

14d87ec5

Add Vulkan tensor copy. (#46481)

Ashkan Aliabadi committed 3 years ago

ec600bc3

Newer Older