strisunshinewentingwang/pytorch

Pull Requests Commits

[vulkan] Adaptive local work group size (#61170)

SS-JIA committed 3 years ago

64ac4288

[vulkan] Calculate a 4x4 output tile for each invocation in conv2d_pw (#60760)

SS-JIA committed 3 years ago

f324421d

[vulkan] Convolution op cleanup (#60759)

SS-JIA committed 3 years ago

a1b5025e

[vulkan] Reduce submission rate to save CPU cycles (#60758)

SS-JIA committed 3 years ago

cacab7e9

[package] merge test_torchscript into test_package_script (#61807)

suo committed 3 years ago

554038c2

ban PyTorchStreamWriter from writing the same file twice (#61805)

suo committed 3 years ago

f02cfcc8

[package] fix storage serialization collision (#61806)

suo committed 3 years ago

04043d68

add BFloat16 support for topk on CPU (#59547)

jiayisunx committed 3 years ago

c30048fc

ignore and clear not ready errors (#61554)

jeffdaily committed 3 years ago

15210f3b

Regenerate libtorch workflow files that got lost in merge conflict (#61872)

janeyx99 committed 3 years ago

e68c0168

Fix grammatical errors on the PyTorch Contribution Guide (#61818)

ndkshr committed 3 years ago

0a6d8824

Port `signbit` to structured kernel (#57936)

Kiyosora committed 3 years ago

43c5dc40

Remote whitespace introduced by #61438 (#61863)

Dmytro Dzhulgakov committed 3 years ago

44d32671

Exclude wrapper tensors from functorch in the native::resize_output fastpath (#61846)

zou3519 committed 3 years ago

26d17ddc

Remove unnecessary Ubuntu version checks (#61738)

xuhdev committed 3 years ago

f9128897

Always use fast gradcheck for LayerNorm 3d_no_affine_large_feature (#61848)

soulitzer committed 3 years ago

1b0a7f38

[BE] Include a unit test for Save Operator with db_options

Kaige Liu committed 3 years ago

094abf5f

Upgrade CPUFallback for loops (#61722)

r-barnes committed 3 years ago

e389650f

[DDP] Add API to get model parameters in hook (#61637)

rohan-varma committed 3 years ago

04bd9d75

Update progress and error reporting in clang-tidy (#61672)

1ntEgr8 committed 3 years ago

66c8d21d

ENH Adds tests and docs for 2d & 3d modules that already support no batch (#61262)

thomasjpfan committed 3 years ago

24a6eb3f

enable check trace when tracing a mkldnn model (#61241)

XiaobingSuper committed 3 years ago

4f46943e

fmin has been ported to the structured kernel, removing the old implementation (#60810)

Freey0 committed 3 years ago

75b68def

fmod: Port to structured (#60809)

Freey0 committed 3 years ago

b526080d

for shared-memory handles, use an atomic counter, instead of potentially colliding random numbers (#60978)

michael committed 3 years ago

b65ddef0

Fix benchmark's import module and remove its usage of tools.stats.scribe (#61808)

zhouzhuojie committed 3 years ago

ac5a40e0

reduce max_num_threads for complex double ops in reduce_kernel (#61438)

Fuzzkatt committed 3 years ago

9c3346c8

Migrate libtorch to GHA (#61774)

janeyx99 committed 3 years ago

d565b3e9

Minor documentation fixes (#61785)

awgu committed 3 years ago

3e3acf8a

Fix indent (#61784)

awgu committed 3 years ago

813b887d

Newer Older