strisunshinewentingwang/pytorch

Pull Requests Commits

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

Tensor class created from java does not call native methods

IvanKobzarev committed 4 years ago

3a19980b

Add test to torch.jit.export_opnames, make the _C function private

Martin Yuan committed 4 years ago

11854bcd

Updating submodules

svcscm committed 4 years ago

81329c90

Exclude lite interpreter Java files from OSS host build

dreiss committed 4 years ago

35b24976

Remove observers in the end (#31407)

jerryzh168 committed 4 years ago

08de70ca

Call `getQSchemeAndQParamMap` later in `quantizeTensors` (#31406)

jerryzh168 committed 4 years ago

b4c48b7e

Use macros instead of directly accessing Python object fields (#31388)

colesbury committed 4 years ago

df9d5b8a

run optimizations on pre-profiled graph (#31392)

Krovatkin committed 4 years ago

5375ceae

Add fake parsing for torchbind classes in schema type parser

James Reed committed 4 years ago

256db1e6

optimize FloatToFused8BitRowwiseQuantized and Fused8BitRowwiseQuantizedToFloat (#31470)

jspark1105 committed 4 years ago

7a12ccd0

Im2col export (#30972)

neginraoof committed 4 years ago

0b57b383

Make fully_qualified_type_name_impl() compatible with VS2017 15.9 (#31455)

malfet committed 4 years ago

6cd987e7

Fix input_channels divisibility check in concat_split_op (#31448)

malfet committed 4 years ago

2099cfa1

Test reading `__cuda_array_interface__` inferred strides. (#31451)

gchanan committed 4 years ago

b38901aa

Newer Older