strisunshinewentingwang/pytorch

Pull Requests Commits

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

CONTRIBUTING.md: add instructions on how to remote desktop into Windows CI (#50841)

janeyx99 committed 3 years ago

49544171

fix typo of quantized README.md (#50681)

XiaobingSuper committed 3 years ago

c945a5bb

Skip test_variant_consistency_eager_addr_cpu_bfloat16 (#50836)

anjali411 committed 3 years ago

7fdc6a27

Use doctest directly to get docstring examples (#50596)

samestep committed 3 years ago

c147aa30

[TensorExpr] Use wider type for scalars (#50774)

asuhan committed 3 years ago

1bde5a21

[pytorch] fix ConstRefCType usage in codegen/api/native.py (#50742)

ljk53 committed 3 years ago

24fd8431

Add support for NCCL alltoall (#44374)

zasdfgbnm committed 3 years ago

44922f26

ZeroRedundancyOptimizer: an implementation of a standalone sharded optimizer wrapper (#46750)

blefaudeux committed 3 years ago

87fb3707

Add cloud-tpu-client to xla CI. (#50823)

Ailing Zhang committed 3 years ago

c3e3e606

Remove gtest_prod.h from TP agent. (#50766)

pritamdamania committed 3 years ago

be7e9845

quantization: Linear + BatchNorm1d fusion (#50748)

vkuzo committed 3 years ago

ac8e90fa

Fix replication_pad for cuda launch configuration (#50565)

xwang233 committed 3 years ago

db86dd8a

Newer Older