strisunshinewentingwang/pytorch

Pull Requests Commits

Reduce set of build/tests which run on PRs. (#20775)

ezyang committed 5 years ago

d03265b4

Use Device instead of Backend in TensorIterator (#20690)

colesbury committed 5 years ago

dee11a92

JIT: Eliminate SumToSize by using Optional Lists (#18697)

t-vi committed 5 years ago

17941f99

Update version strings to 1.2

gchanan committed 5 years ago

47043220

Add llvm8 installation step. (#20879)

Roman Dzhabarov committed 5 years ago

a640c815

Update Refinement Docs (#20912)

Elias Ellison committed 5 years ago

fa203276

Remove extra workspace queries in matrix inverse computation (#20904)

vishwakftw committed 5 years ago

8c4b2a83

In Dockerfile, do not install unecessary packages, use conda to install ninja (saving one layer), and use "." to refer to WORKDIR to reduce redundancy. (#20881)

xuhdev committed 5 years ago

4109ec12

Leave it as an option for whether to colorize output during build (#20771)

xuhdev committed 5 years ago

6af24826

tensor_illustration with correct numbers and better fonts for README file (#20751)

PgLoLo committed 5 years ago

ec45baf4

Raise TypeError when the argument to isinf and isfinite is not a tensor (#20817)

xuhdev committed 5 years ago

ef1fdc27

Fix documentation for attention mask shape (#20850)

jotsif committed 5 years ago

87040af4

Use "length of the RNN input" instead of "length of the RNN"

kit1980 committed 5 years ago

a5c90aaf

Instructions for how to update pytorch-ci-hud when updating binary builds (#20758)

ezyang committed 5 years ago

3e4f213e

Resend "Split ATen/Parallel into interface and backend" (#20825)

Ilia Cherniavskii committed 5 years ago

c3d05e86

Fix init_thread calls in thread pool initialization (#20848)

Ilia Cherniavskii committed 5 years ago

6b748567

Change the quantizer to match the behavior of the FBGEMM implementation (#20892)

Zafar Takhirov committed 5 years ago

1bb728fe

Catchall kernels instead of fallback kernels (#20773)

smessmer committed 5 years ago

fc941d3b

Lightweight at-most-once logging for API usage (#20745)

Dmytro Dzhulgakov committed 5 years ago

c25e3378

Remove Variable::Impl and DifferentiableViewImpl (#17072)

Will Feng committed 5 years ago

8cde4c4d

Adding ShufflenetV2 to caffe2's benchmark suite. (#20180)

Abhinav Jauhri committed 5 years ago

f93e0619

Updating submodules

svcscm committed 5 years ago

3aa7ee6f

Updating submodules

svcscm committed 5 years ago

cfb6c4a8

dropout symbolic_script should respect the training flag (#20760)

suo committed 5 years ago

62af37aa

Move torchvision install out of onnx test script

bddppq committed 5 years ago

bd53c8eb

Dict is a reference type (#20669)

smessmer committed 5 years ago

d5b7138a

bug fix 19374 - fix for upsample export

pk-g committed 5 years ago

93d5503f

Fix oscillation in coalesceInsertedDataDependencies (#20833)

Yinghai Lu committed 5 years ago

48bf7b9b

Use conda torchvision version (#20865)

Will Feng committed 5 years ago

2d96876d

Move THCTensor_{random, clampedRandom, cappedRandom} to ATen (#20620)

syed-ahmed committed 5 years ago

b6d0f6c8

Newer Older