strisunshinewentingwang/pytorch

Pull Requests Commits

Add OpInfo for `nn.functional.cosine_embedding_loss` (#67465)

ejguan committed 2 years ago

bf319951

Add OpInfor for `nn.functional.ctc_loss` (#67464)

ejguan committed 2 years ago

bcd301a4

Add OpInfo for `nn.functional.poisson_nll_loss` (#67371)

ejguan committed 2 years ago

e2e20e79

Add OpInfo for `nn.functional.gaussian_nll_loss` (#67376)

ejguan committed 2 years ago

8b8fb4f4

Add OpInfo for `nn.functional.hinge_embedding_loss` (#67381)

ejguan committed 2 years ago

1d900ee2

Add OpInfo for `torch.nn.functional.gaussian_nll_loss` (#67356)

ejguan committed 2 years ago

c6a6c093

Sort output of *NativeFunctions.h (#67046)

silvasean committed 2 years ago

2e156f64

Norms Op Info (#67442)

Samantha Andow committed 2 years ago

f95ed474

add dequantize support for fp16 + cuda (#67234)

vkuzo committed 2 years ago

d58f2093

pytorch quantization: document the custom module APIs (#67449)

vkuzo committed 2 years ago

99282126

[quant][graphmode][fx] Add support for ObservationType.OUTPUT_SHARE_OBSERVE_WITH_INPUT in backend_config_dict (#67210)

jerryzh168 committed 2 years ago

acdc7549

[quant] Move test file to fx2trt folder (#67129)

jerryzh168 committed 2 years ago

2bb20c0e

Fixes to make trt timing_cache really work (#67524)

Yinghai Lu committed 2 years ago

5e46a4f6

[deploy] fix TypedStorage serialization (#67499)

suo committed 2 years ago

96c86821

Revert D31450501: Wextra caffe2/

Rob Kinyon committed 2 years ago

4052393a

Fix Ads build broken due to comparison type mismatch (#67526)

Vidhoon Viswanathan committed 2 years ago

18807273

Remove the argument strip_doc_string of export() method entirely. (#66615) (#67278)

fatcat-z committed 2 years ago

26241994

Deprecate the argument _retain_param_name of export() method entirely. (#66617) (#67277)

fatcat-z committed 2 years ago

43d51254

[ONNX] Remove the argument enable_onnx_checker of export() method entirely. (#66611) (#67276)

fatcat-z committed 2 years ago

40920185

[ONNX] Update value name copying logic for onnx (#66170) (#67275)

BowenBao committed 2 years ago

609da981

Wextra caffe2/ (#67319)

r-barnes committed 2 years ago

7c2d3e6d

Workaround the bug of TRT which creates extra outputs (#67327)

Yinghai Lu committed 2 years ago

d8bde98f

Add Initial NNC Dynamic Shapes Flow (#66136)

Elias Ellison committed 2 years ago

fc82ad18

Adding support for Symbolic Shapes in Inplace Ops #65642 (#65729)

Gamrix committed 2 years ago

26615074

ci: Migrate hardcoded docker builds to GHA (#67455)

seemethere committed 2 years ago

d0bc01fa

Adding `optimizer` attribute to SequentialLR (#67406)

yiwen-song committed 2 years ago

6696c59a

[SR] Native implementation for aten::size (#67346)

Mike Iovine committed 2 years ago

354363b5

[PyTorch][easy] Deduplicate memory planner creation code (#67265)

swolchok committed 2 years ago

9f01937c

Revert D31894777: [pytorch][PR] Replace issue templates with new issue forms

NicolasHug committed 2 years ago

82c35650

[SR] Native implementation for aten::view (#67341)

Mike Iovine committed 2 years ago

afb84344

Newer Older