strisunshinewentingwang/pytorch

Pull Requests Commits

Add temporary torch::k{name} enum declarations

Will Feng committed 4 years ago

e20d5c28

Add boxed fallback (#26368)

ezyang committed 4 years ago

7c465aae

make cudnn rnn respect current stream (#27026)

ngimel committed 4 years ago

2bcbc150

Integrate FC fp16 exporter into Dper2 (#26582)

Xing Wang committed 4 years ago

a1513dce

C++ API parity: AdaptiveAvgPool3d

pbelevich committed 4 years ago

1a3997e0

Improve repr for quantized modules

James Reed committed 4 years ago

4d7bec5f

Migrate le/gt/ge/eq/ne from the TH to Aten. Added support of type promotion. (#27017)

ifedan committed 4 years ago

ee2c79d6

Turn Caffe2 CUDA 9.1 + py2 to CUDA 10.1 + py3 (#26835)

houseroad committed 4 years ago

9e1cf99a

Per-channel baseline (#26516)

raghuramank100 committed 4 years ago

2ccbdb79

C++ API parity: AdaptiveAvgPool2d

pbelevich committed 4 years ago

a31fd5ea

C++ API parity: AdaptiveAvgPool1d

pbelevich committed 4 years ago

7d58060f

move parallel_for/parallel_reduce common implementation to cpp (#26969)

ljk53 committed 4 years ago

18dd541b

Resubmit [pytorch][PR] [ONNX] Updating producer_version in exported O… (#27004)

houseroad committed 4 years ago

a173bea4

Add InsertPackUnpack pass (#26959)

jerryzh168 committed 4 years ago

9f9ba3a9

Fix race condition in Function::optimized_graph(). (#27012)

Mikhail Zolotukhin committed 4 years ago

310ebb45

Revert D17635651: [pytorch][PR] Migrate le/gt/ge/eq/ne from the TH to Aten. Added support of type promotion.

Mike Ruberry committed 4 years ago

8858f42a

Add std::variant backport as c10::variant (#26836)

Will Feng committed 4 years ago

0cd18803

Dont zero out buffers in dynamic linear (#27002)

James Reed committed 4 years ago

cca3a369

PyTorch Graph Mode Quantization API (#26390)

jerryzh168 committed 4 years ago

09f0e949

Fix race condition in torch::jit::Function (#27009)

Dmytro Dzhulgakov committed 4 years ago

da93cc5c

Remove unimplmented passes (#26978)

jerryzh168 committed 4 years ago

f8db764f

Move patterns in QuantFusion to a separate file (#26848)

jerryzh168 committed 4 years ago

76676765

Move some class/functions in test_jit.py to jit_utils.py (#26839)

jerryzh168 committed 4 years ago

5e79b5b1

Add QuantFusion to graph_executor (#26591)

jerryzh168 committed 4 years ago

b0f1b5c7

Migrate le/gt/ge/eq/ne from the TH to Aten. Added support of type promotion. (#26981)

ifedan committed 4 years ago

541de7e1

fix range for non-int inputs and pow implementation (#26926)

Elias Ellison committed 4 years ago

ff8b7ef6

Rewrite argmax and argmin as TensorIterator reductions (#26181)

peterbell10 committed 4 years ago

9080f1c5

Add torch.promote_types function

nairbv committed 4 years ago

0c6a18de

Add fakefp16 transformation.

ipiszy committed 4 years ago

024a422f

Add optimized quantize function for ARM (#26867)

supriyar committed 4 years ago

aa0b2842

Older