onnxruntime/onnxruntime/core/codegen/passes
KeDengMS 60208463a9
[NupharEP] Enable parallel schedule (#2505)
* [NupharEP] Enable parallel schedule
* Update TVM with the fix to TVM threadpool to use OpenMP if possible
* Add parallel schedule when trying to vectorize
With this change, BERT squad perf on a 4-core (8 HT) CPU goes from 187ms to 150ms

* Address CR, docs and cmake update

* Doc fix

* Fix mkl

* Fix TVM windows build when using mklml
2019-11-28 08:35:56 -08:00
..
op_ir_creator [NupharEP] Multiple optimizations (#2380) 2019-11-14 10:40:33 -08:00
scheduler [NupharEP] Enable parallel schedule (#2505) 2019-11-28 08:35:56 -08:00
utils Introduce PrimitiveType into a Type System along with an integer constant (#2307) 2019-11-08 17:47:06 -08:00
weight_layout Nuphar: Fix a bug in weight layout where read may go out of bound (#2129) 2019-10-15 00:11:41 -07:00