strisunshinewentingwang/pytorch

Pull Requests Commits

refactor caffe2 operator constructors - 9/9 (#17090)

smessmer committed 5 years ago

a9395ce2

Fix the false generated_comment (#17563)

gemfield committed 5 years ago

9bcceb75

Remove useless OpenCV reference

Dmytro Dzhulgakov committed 5 years ago

13e6326c

convolution/matmul/dropout (#17523)

Ailing Zhang committed 5 years ago

03132c1f

disallow shape analysis with resize ops (#17518)

Elias Ellison committed 5 years ago

221edddd

Make C10_MOBILE consistent with how feature macros are usually used (#17481)

smessmer committed 5 years ago

6706e9af

Disable c10 dispatcher on mobile (#17078)

smessmer committed 5 years ago

7c5ffc41

Always synchronize src and dst streams when copying tensors (#16966)

mrshenli committed 5 years ago

11545065

ONNX Export Adaptive Pooling

lara-hdr committed 5 years ago

5f06dcc4

Use name for output variables instead of out in JIT (#17386)

cpuhrsch committed 5 years ago

e47aeede

Forcing UTC on Mac circleci jobs (#17516)

pjh5 committed 5 years ago

1971c052

Fix math::Set for large tensor (#17539)

xiaomengy committed 5 years ago

9709d5e7

Add sparse gradient option to `gather` operation (#17182)

Natalia Gimelshein committed 5 years ago

b4572668

add elastic zeus handler (#16746)

janewangfb committed 5 years ago

a2b9f7f4

optimize elementwise sum (#17456)

jspark1105 committed 5 years ago

222a0786

Enable boolean_mask, adadelta, adagrad fp16 on ROCm (#17235)

rohithkrn committed 5 years ago

8c722178

Enabled HALF for fill() and zero() methods. Moved them into THTensorFill (#17536)

izdeby committed 5 years ago

e0b44cac

Fix autograd with buffers requiring grad in DataParallel (#13352)

ssnl committed 5 years ago

44a607b9

enable assymetric dilations and stride for miopen conv (#17472)

lcskrishna committed 5 years ago

74098ead

Enable tests working on ROCm 2.1 dual gfx906

iotamudelta committed 5 years ago

76828647

Fix linking errors when building dataloader test binaries on Windows (#17494)

peterjc123 committed 5 years ago

96faaa9d

hysts committed 5 years ago

cbefd032

Remove Bool/IndexTensor from schema for native functions with derivatives (#17193)

cpuhrsch committed 5 years ago

eff672ef

Fix operator initialization order (#15445)

Ilia Cherniavskii committed 5 years ago

348d1889

Make transpose consistent with numpy's behavior (#17462)

bhushan23 committed 5 years ago

4ca1a545

Bump up the ONNX default opset version to 10 (#17419)

houseroad committed 5 years ago

63519df0

' ' ==> ' ' (#17498)

liangdzou committed 5 years ago

72eb70c2

Support all ROCm supported uarchs simultaneously: gfx803, gfx900, gfx906 (#17367)

iotamudelta committed 5 years ago

1607bb32

refactor: a bit intricate so I refactor it (#16995)

knightXun committed 5 years ago

5903522a

new batch of expect file removals

Elias Ellison committed 5 years ago

e5b4baab

Newer Older