strisunshinewentingwang/pytorch

Pull Requests Commits

try liblapack_static, may need `if cuda >= 10.1`

xwang233 committed 3 years ago

3075adc8

Merge remote-tracking branch 'upstream/master' into cusolver-svd

xwang233 committed 3 years ago

a99df7d9

[FX] Update overview docstring (#50896)

James Reed committed 3 years ago

50166379

[distributed_test]Enable disabled ROCm tests. (#50421)

jaglinux committed 3 years ago

eb0fe706

[static runtime] Shortcut resize_({0})

bertmaher committed 3 years ago

aa3c28a2

install magma for cuda 11.2 in conda (#50559)

janeyx99 committed 3 years ago

8e9ed27a

[ONNX] Handle sequence output for models (#50599)

neginraoof committed 3 years ago

137f2a38

Add autograd tests for complex matrix norm nuclear and +/-2 (#50746)

kurtamohler committed 3 years ago

c082e218

Automated submodule update: tensorpipe (#50895)

facebook-github-bot committed 3 years ago

201f0c1f

add and adjust kernel launch checks under fbcode/caffe2/caffe2/utils (#50862)

Dan Fan committed 3 years ago

3cd8ed97

Add batched grad testing to OpInfo (#50818)

zou3519 committed 3 years ago

16691516

Update Kineto revision (#50855)

Ilia Cherniavskii committed 3 years ago

1cce4c5e

Miscellaneous batched grad testing (#50738)

zou3519 committed 3 years ago

884fb487

[te] Speed up relu on cpu

bertmaher committed 3 years ago

8ede828d

[android] Fix YUV camera image to tensor (#50871)

IvanKobzarev committed 3 years ago

98e29146

[quant][doc] Adding a table comparing eager and fx graph mode (#50413)

jerryzh168 committed 3 years ago

b5242d66

Revert D25976245: [pytorch][PR] Enable Skipped ROCM Tests in common_nn.py

ngimel committed 3 years ago

4d169258

Adds per-op microbenchmarks for NNC (#50845)

Chillee committed 3 years ago

4cca0836

Improve call provenance during GraphModule scripting (#50538)

Ansley Ussery committed 3 years ago

4ac48909

[optimizer] refactor AdamW to use functional API (#50411)

wanchaol committed 3 years ago

df963449

[optimizer] refactor RMSProp to use functional API (#50410)

wanchaol committed 3 years ago

ce1781d8

[optimizer] refactor Adadelta to use functional API (#50409)

wanchaol committed 3 years ago

d6fb27ce

[optimizer] refactor SGD to use functional API (#45597)

wanchaol committed 3 years ago

a0cf5566

Add CPP Full Reduction Benchmarks. (#50193)

zheng-xq committed 3 years ago

b96a6516

Add full reduction benchmark. (#50057)

zheng-xq committed 3 years ago

88b36230

Enable Skipped ROCM Tests in common_nn.py (#50753)

arindamroy-eng committed 3 years ago

24a02721

Automated submodule update: tensorpipe (#50807)

facebook-github-bot committed 3 years ago

480bb7d3

[Gradient Compression] Fix warm-start for PowerSGD laywerwise compression (#50283)

Yi Wang committed 3 years ago

439afda0

[FX][docs] Add limitations of symbolic tracing (#50638)

James Reed committed 3 years ago

d0e942f9

Make `split_module` results deterministic (#50470)

Ansley Ussery committed 3 years ago

c88eed97

Older