strisunshinewentingwang/pytorch

Pull Requests Commits

Adding quantized clamp kernel (#30541)

dskhudia committed 4 years ago

a2463cbc

Update ONNX Flatten to accept negative indices in opset 11 (#30751)

lara-hdr committed 4 years ago

1d5af959

move AWS ECR gc jobs to circleci (#30996)

Mingbo Wan committed 4 years ago

84d67966

fix torch_train build (#30497)

linbinyu committed 4 years ago

5c936845

Only create OwnerRRefs when processing remote calls (#31163)

mrshenli committed 4 years ago

a38184db

Enabled roll for bool tensor (#31194)

izdeby committed 4 years ago

f6c31f61

remove / rewrite weak module tests (#31193)

Elias Ellison committed 4 years ago

bee6344d

Re-apply "[bert/RoBERTa] Optimize LayerNorm with explicit vectorization using Vec256" (#31127)

jianyuh committed 4 years ago

066e3ed9

Adding function to convert Module to channels last

VitalyFedyunin committed 4 years ago

66f2bba8

Fix CircleCI behavior for non-leaf stack PRs (#31088)

IvanKobzarev committed 4 years ago

4ead2e89

Remove unnecessary ATen/core/EnableNamedTensor.h (#31117)

zou3519 committed 4 years ago

bcb0bb7e

Remove all remaining usages of BUILD_NAMEDTENSOR (#31116)

zou3519 committed 4 years ago

9047d4df

Revert D18923167: Expose setNumThreads to android api

suo committed 4 years ago

c0bcfd04

Resubmit overload v2 (#31123)

Elias Ellison committed 4 years ago

56de8853

Remove `insert_prepack_unpack` and `fold_prepack` for now (#30909)

jerryzh168 committed 4 years ago

3a02ed82

Add types for the remaining optimizers. (#31130)

stephenroller committed 4 years ago

159835e6

Tweak pollTimedOutRPCs thread synchronization (#30355)

xush6528 committed 4 years ago

2488231f

Re-enable python 2 builds (#31164)

suo committed 4 years ago

0db6c013

Add native/quantized to the list of header rewrites (#31151)

Serhat Yilmaz committed 4 years ago

4f5a4be4

Partially support tensor lists in loop/concat/stack (#30126)

BowenBao committed 4 years ago

6ab2d1b1

Change type of timeoutFutures_ key to time_point instead of duration (#31078)

xush6528 committed 4 years ago

a3ed350e

Make Conv{1,2,3}dOptions and ConvTranspose{1,2,3}dOptions different classes (#31005)

Will Feng committed 4 years ago

49a5841a

Fix type unification With Specialized Tensor Shapes (#31076)

Elias Ellison committed 4 years ago

85107e72

ONNX Interpolate Add Scales Params (#28324)

lara-hdr committed 4 years ago

97c1e90f

Add ONNX Export Support to floor_divide (#31081)

lara-hdr committed 4 years ago

79c27ba4

Updating submodules

svcscm committed 4 years ago

d81c6bde

dynamicly quantized linear benchmarking

z-a-f committed 4 years ago

efe683fb

Make rref fetch calls async. (#31086)

jjlilley committed 4 years ago

73f9e816

Unify list elements for all list types (#30777)

davidriazati committed 4 years ago

679b20b1

doc fix for max method: a warning about different behaviour on CPU and GPU (#31115)

nikitaved committed 4 years ago

04144630

Newer Older