strisunshinewentingwang/pytorch

Pull Requests Commits

Fix deviceCount on FakeGuardImpl. (#18745)

ezyang committed 5 years ago

52a3a514

Stop swapping in Storages of the wrong device for Tensors. (#18831)

gchanan committed 5 years ago

486fae56

Pass ScalarType separately from Type in python constructors

Roy Li committed 5 years ago

d70c6f23

Store ScalarType and Backend instead of Type in TensorIterator

Roy Li committed 5 years ago

f5741eb8

Introduce DeprecatedTypeProperties class (#17991)

Roy Li committed 5 years ago

c705d9eb

Fix to handle null strides in DLPack tensor (#18510)

bwasti committed 5 years ago

095f88e0

Add shape inference function for Split (#18838)

Yinghai Lu committed 5 years ago

e5e2110a

Fix the duplication problem in _unique_state_dict (#18139)

houseroad committed 5 years ago

0c237f13

fold col offset into bias; optimize A symmetric quant (#17026)

jspark1105 committed 5 years ago

fa0ad057

fix flake8 lint (#18835)

suo committed 5 years ago

72913a55

run cpp tests for non-cuda builds in test_jit.py (#18826)

suo committed 5 years ago

0a4117a3

Fix the linter (#18842)

houseroad committed 5 years ago

100f95a3

Enforce single parent for script submodules (#18379)

zdevito committed 5 years ago

7e59c604

Allow ints, floats, and tensors in conditional (#18755)

Elias Ellison committed 5 years ago

b80a4fa2

Fix layernorm ad formula on weight and bias (#18233)

wanchaol committed 5 years ago

843e6234

Unify namespace of script::Module (#18378)

zdevito committed 5 years ago

0512e4e3

Step 1: Secretly add return_counts to unique, and refactor unique_dim for performance (#18648)

VitalyFedyunin committed 5 years ago

773ce4fb

Support replicating multi-GPU modules (#18687)

mrshenli committed 5 years ago

7ae0263e

wanchaol committed 5 years ago

eabd9eac

Remove `device_guard: False` from native_functions that don't have a … (#18803)

gchanan committed 5 years ago

862aff64

Switch our Linux machine AMI to a newer image. (#18433)

ezyang committed 5 years ago

cb959aa7

QTensor (#18230)

jerryzh168 committed 5 years ago

dfcd7b01

Enforce import order to make protobuf cpp implementation in python work (#18560)

Dmytro Dzhulgakov committed 5 years ago

3af2d6d9

Pin onnx ir_version to 4 (#18768)

houseroad committed 5 years ago

3b71f2e1

fix nccl compilation to make sure it compiles for architectures that pytorch compiles for (#18739)

soumith committed 5 years ago

8711df89

push magma init into lazyInitCUDA (#18527)

soumith committed 5 years ago

b5d8844b

For some files that are touched by the QTensor diff (#18765)

jerryzh168 committed 5 years ago

ed9724f3

Fix contiguous AD and Autogradzero inconsistency (#18633)

wanchaol committed 5 years ago

a21e256e

Added indexing for bool tensors and bool Indices (#18583)

izdeby committed 5 years ago

5950c1e8

add an assertion to check the param num (#18145)

houseroad committed 5 years ago

65dfe120

Newer Older