strisunshinewentingwang/pytorch

Pull Requests Commits

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

add type promotion support for sparse tensors (#30429)

nairbv committed 4 years ago

d0d6e0b5

Updating submodules

svcscm committed 4 years ago

e9ef087d

modify model to enable loading by blob (#31507)

ChunliF committed 4 years ago

4c341582

Add support for `del` (#31273)

davidriazati committed 4 years ago

06dbef66

Don't dispatch to cudnn if it is not possible to make it 32bit by splitting batch dim (#31383)

zasdfgbnm committed 4 years ago

624088e4

Updating submodules

svcscm committed 4 years ago

87768e5a

Newer Older