strisunshinewentingwang/pytorch

Pull Requests Commits

Automatic update of fbcode/onnx to 57ebc587fcf3913b4be93653b0dd58c686447298 (#31642)

houseroad committed 4 years ago

204939b4

Clean White List for BC Checks (#31629)

houseroad committed 4 years ago

ffcac9ad

Integrating MaskedAdagrad

chocjy committed 4 years ago

4983ef8d

cudnn grouped convolution nhwc patch (#31444)

jjsjann123 committed 4 years ago

909b8eba

Create byte-aware word lstm benchmark (#31260)

Fan Wang committed 4 years ago

39508501

avinashsai committed 4 years ago

91eb7c26

Updating submodules

svcscm committed 4 years ago

34dce8e3

Add Python language reference docs (#30686)

davidriazati committed 4 years ago

ec4e3477

Print all broken ops instead of the first one (#31628)

houseroad committed 4 years ago

5d95a9ca

Updating submodules

svcscm committed 4 years ago

cf46bcac

Ensure legacy sparse constructor/new doesn't interpret python data as tensor data. (#31490)

gchanan committed 4 years ago

866c1b1f

Updating submodules

svcscm committed 4 years ago

e2951d58

Error out if legacy Tensor.new is called on alternate layouts / dtypes (#31485)

gchanan committed 4 years ago

29f34583

revert D18805532 and make numerics of masked adagrad consistent with unmasked adagrad (#30784)

jspark1105 committed 4 years ago

a54dc87e

Bypass _TorchScriptTesting_StackString::pop in BC check now (#31586)

houseroad committed 4 years ago

363d8be7

Fix null pointer dereference on Android for strtod_c (#31582)

Kaikai Wang committed 4 years ago

46ad80c8

Fix parsing of big float literals (#29940)

davidriazati committed 4 years ago

446e9af5

Conv transpose/backward split 32bit (#31510)

zasdfgbnm committed 4 years ago

218cfd56

Remove -Wno-unused-private-field

aditya7fb committed 4 years ago

fb63c0e2

Support optional float parameters (float?, optional<double>). (#31517)

gchanan committed 4 years ago

68e51723

Raise warning for schedulers following chainable shedulers (#31125)

vincentqb committed 4 years ago

9459db86

fix test_process_group_debug_info flaky test (#31533)

rohan-varma committed 4 years ago

fe76af96

add enabled API to autograd profiler (#31380)

rohan-varma committed 4 years ago

cc2d5ca3

Separate torchbind from Python (#30242)

James Reed committed 4 years ago

7d630278

set stream everytime when we get a cuDNN handle (#31541)

zasdfgbnm committed 4 years ago

700109eb

set stream everytime when we get a cuSparse handle (#31538)

zasdfgbnm committed 4 years ago

b5bbec7b

set stream everytime when we get a cuBlas handle (#31537)

zasdfgbnm committed 4 years ago

8d8e8288

Split on batch dimension when 32bit indexing not enough for convolution forward (#31379)

zasdfgbnm committed 4 years ago

0b0f90f5

make gc script python2 compatible (#31536)

Mingbo Wan committed 4 years ago

3820d6f6

Nightly dimension, input shape in gradle (#30195)

IvanKobzarev committed 4 years ago

c808eed0

Newer Older