strisunshinewentingwang/pytorch

Pull Requests Commits

ns for fx: qat test cases for unshadowed activations (#55508)

vkuzo committed 3 years ago

37fbc069

ns for fx: extend functional weight extraction testing to QAT (#55507)

vkuzo committed 3 years ago

f6a3936a

ns for fx: add qat handling for weight extraction (#55506)

vkuzo committed 3 years ago

1cbc4023

ns for fx: make NSTracer inherit from QuantizationTracer (#55505)

vkuzo committed 3 years ago

3786c271

ns for fx: change node I/O determination to strict allowlist (#55434)

vkuzo committed 3 years ago

5ad3bc71

[kineto] deprecate metdata args from ClientTraceActivity (#55988)

chaekit committed 3 years ago

1ca51f0f

Python API for Vitals (#53238)

Victor Bittorf committed 3 years ago

52f1a07b

Port all unary float functions to structured (#56082)

ezyang committed 3 years ago

f17c9ea2

Change all unary functions stubs to use TensorIteratorBase& (#56078)

ezyang committed 3 years ago

cfc97162

remove annoying warnings from common_nn.py (#55982)

Chillee committed 3 years ago

3c4e1cd1

Add cost inference for MulGradient operator

Chao Kong committed 3 years ago

ff1498e6

port addmv to structured kernels (#55746)

ngimel committed 3 years ago

3fbca31b

Reland: D27652485: [nnc] Enable CPU fusion only when num_threads == 1" (#56120)

bertmaher committed 3 years ago

8e82e932

[reland][ROCm] use hiprtc precompiled header (#55965)

jeffdaily committed 3 years ago

e1752ffa

Fix ChanelShuffle named tensor warnings (#55911)

h6197627 committed 3 years ago

f02454f9

[dist_optim] add distributed functional rprop optimizer (#55834)

wanchaol committed 3 years ago

dd090e72

[dist_optim] Add distributed functional Adamax optimizer (#55833)

wanchaol committed 3 years ago

4e9e7200

[optim] refactor rprop to use functional API (#55832)

wanchaol committed 3 years ago

8ef13cf9

[optim] refactor adamax to use functional API (#55830)

wanchaol committed 3 years ago

bb245b64

[quantization] Fix deepcopy on quantized ConvNd (#56154)

James Reed committed 3 years ago

f26a6cb3

Add complex autograd support for `torch.cross` (#55854)

kurtamohler committed 3 years ago

a3a75bd3

Revert D27753803: [19/n][torch/elastic][upstream] Replace pytorch.distributed.launch with torchelastic launcher

VitalyFedyunin committed 3 years ago

90e103dd

[torch/elastic] Introduce `PeriodicTimer` (#55919)

cbalioglu committed 3 years ago

512c744f

Revert D27758303: [20/n][torch/elastic][upstream] Move torchelastic.distributed.tests to pytorch.distributed

VitalyFedyunin committed 3 years ago

e2036ea3

should_check_autodiff is now should_autodiff_node (#56013)

mattip committed 3 years ago

9bfe16a3

[caffe2] allow passing options to the DB in Save operations (#55935)

simpkins committed 3 years ago

aae1023b

Add support for refinement for torch.jit.Future (#56148)

nikithamalgi committed 3 years ago

14d529a3

Revert "Deprecate legacy constructor `torch.Tensor()` (#54414)" (#55831)

heitorschueroff committed 3 years ago

33159b68

[nnc] Don't fuse fp16 on CPU (#56119)

bertmaher committed 3 years ago

b9405160

[nnc][trivial] Trailing underscore style for llvmCode, asmCode members (#56118)

bertmaher committed 3 years ago

16820bba

Newer Older