strisunshinewentingwang/pytorch

Pull Requests Commits

Export ONNX Dropout for opset 10 (#20710)

lara-hdr committed 5 years ago

27d1daab

math module support: isnan, asinh, atanh, cosh, sinh, and tanh (#19337)

lhendre committed 5 years ago

770089c2

Remove onnx export expects (#21238)

Elias Ellison committed 5 years ago

fb726252

add contiguous function type hint for tensor (#21285)

shihongzhi committed 5 years ago

2e59a0a6

Write add_scalars data in the same file (#21100)

nataliakliushkina committed 5 years ago

96667dfe

Fix build error in c10 on Windows (#21005)

peterjc123 committed 5 years ago

5b336987

Move THCTensor_{normal, normal_means, normal_stddevs, normal_means_stddevs} to ATen (#21287)

syed-ahmed committed 5 years ago

155f7673

EliminateGuards

Krovatkin committed 5 years ago

21113c2d

Make is_contiguous checks generic in number of arguments (#21106)

colesbury committed 5 years ago

c8539be9

Significantly simplify the spawning of pytorch libs building process. (#21105)

xuhdev committed 5 years ago

b159e0ce

Retry Fix Python DataParallel RNN in no_grad mode (#21262)

mrshenli committed 5 years ago

f62a0060

Fix gelu documents (#21265)

xiaomengy committed 5 years ago

0c6efbd4

Add autograd for layer_norm on CPU (#20883)

xiaomengy committed 5 years ago

eaa3ba65

Add gelu gradient for pytorch (#21237)

xiaomengy committed 5 years ago

31c79b71

Add gelu activation in pytorch (#20665)

xiaomengy committed 5 years ago

93ae040f

Revert D15577342: [pytorch][PR] Fix Python DataParallel RNN in no_grad mode

Karl Ostmo committed 5 years ago

aac424a6

Fixes a bug in the test (#21146)

Zafar Takhirov committed 5 years ago

360e6d1b

Exclude file:line from graphs used for fuser kernel cache (#21252)

James Reed committed 5 years ago

62ae348d

Save the weight shape info the first time we have chance to extract it (#21233)

Yinghai Lu committed 5 years ago

7c40576c

Revert D15548138: Export feature length information for onnxifi operator

Jiang Wu committed 5 years ago

0efc527d

Fix Python DataParallel RNN in no_grad mode (#21197)

mrshenli committed 5 years ago

51ebbe97

Fix typo in test_dataloader

ssnl committed 5 years ago

f051fbd4

compare scalar device with common device (#21236)

Natalia Gimelshein committed 5 years ago

d168a853

Fix wrong type hints for Tensor.is_cuda, is_leaf (#21192)

sublee committed 5 years ago

41b17e24

Fix `sccache not being used on Windows` (#21248)

peterjc123 committed 5 years ago

be7fc406

Add file-line info for jit.load and string frontend (#21217)

James Reed committed 5 years ago

619261d7

Lazily build error strings in schema matching using replay. (#21241)

Owen Anderson committed 5 years ago

b663eec1

fix contribution and governance links (#21243)

sovvo committed 5 years ago

5bc7c1f8

Export feature length information for onnxifi operator (#21110)

ChunliF committed 5 years ago

85786bea

add PT maxpool and avgpool ops to the benchmark suite (#21200)

mingzhe09088 committed 5 years ago

516ea33f

Newer Older