strisunshinewentingwang/pytorch

Pull Requests Commits

Add nvidia-smi memory and utilization as native Python API (#69104)

vincentpierre committed 2 years ago

1365301b

OpInfo for `nn.functional.dropout2d`, revise sample inputs for `dropout` (#67891)

krshrimali committed 2 years ago

51ac60da

add `OpInfo` for `torch.diagflat` (#65680)

pmeier committed 2 years ago

fd5f6b29

add `OpInfo` for `torch.nn.functional.kl_div` (#65469)

pmeier committed 2 years ago

2ba3f8a1

add python dispatch test back to CI and fix typo in test (#69565)

albanD committed 2 years ago

cbd2e499

[Static Runtime] Avoid evaluating expressions of `Node*` for interpreter fallback op (#69489)

d1jang committed 2 years ago

24010abb

Updating quantization documentation (#68907)

HDCharles committed 2 years ago

5ace9c22

[DataLoader] Implementing communication processes for Map-style DataPipes (#68549)

NivekT committed 2 years ago

76d4b4fb

[quant][embdding qat] Re-land Add FX support for QAT EmbeddingBag (#69334)

b-koopman committed 2 years ago

136af3c6

[quant][embedding qat] Re-Land Support Embedding QAT via FX API (#69333)

b-koopman committed 2 years ago

aa026a9e

histc: Avoid dispatch in parallel region (#68520)

peterbell10 committed 2 years ago

043cc65b

Fix TSAN issue in TCPStore (#69590)

pritamdamania committed 2 years ago

2eac4913

Release GIL during RPC shutdown. (#69586)

pritamdamania committed 2 years ago

b1ef941b

OpInfo tests for `(svd|pca)_lowrank` (#69107)

nikitaved committed 2 years ago

ab4ec36b

Fix sign op converter (#69580)

Shirong Wu committed 2 years ago

1b2f68b5

Remove finput and fgrad_input from slow3d transpose signatures (#68899)

jbschlosser committed 2 years ago

cfcb2a90

[quant][graphmode] initial support for fusion pattern in backend_config_dict (#69335)

jerryzh168 committed 2 years ago

84132e3f

use irange for loops (#69533)

r-barnes committed 2 years ago

d6e54939

[Reland][Autograd/Checkpoint] Checkpoint implementation without reentrant autograd (#69508)

rohan-varma committed 2 years ago

8464f072

Allow build_android.sh to forward Vulkan args (#69332)

SS-JIA committed 2 years ago

1bd4ed4f

[vulkan] Disable shader optimization to avoid Validation Errors (#69331)

SS-JIA committed 2 years ago

984fb0c8

[vulkan] Add image format qualifier to glsl files (#69330)

SS-JIA committed 2 years ago

8516f930

[quant][graphmode] Refactor fusion to use the new Pattern format (#68770)

jerryzh168 committed 2 years ago

73dc9233

use irange for loops 6 (#66742)

r-barnes committed 2 years ago

0228dcf4

CMake: Support dynamic codegen outputs (#68246)

peterbell10 committed 2 years ago

5199581a

Rationalize API exports in torch_python (#68095)

peterbell10 committed 2 years ago

a4613208

Output UnionType str rep with () instead of [] (#69502)

tugsbayasgalan committed 2 years ago

c4a780c9

Sparse CSR CUDA: Add block torch.addmv when mat is sparse (#68708)

IvanYashchuk committed 2 years ago

22acb85d

skip ORT tensor in TensorIterator because it doesn't have storage (#68705)

souptc committed 2 years ago

df2ed71b

[SR] Simplify memory re-use algorithm (#68302)

Mike Iovine committed 2 years ago

dffa3ab8

Newer Older