Jing Pu 54f4522a5c Specalize f32->i8/u8 Quanitization with C++ native arithmetic to optimize performance. ...

The CL adds a rounding mode flag to the class and changes the default to rmNearestTiesToAway from rmNearestTiesToEven because 1) Tensorflow QuantizeV2 ops uses rmNearestTiesToAway; 2) the specialization only implements rmNearestTiesToAway.

PiperOrigin-RevId: 270600739

2019-09-22 22:07:51 -07:00

..

AffineOps

Add utility to extract strides from layout map in MemRefType.

2019-09-20 09:26:21 -07:00

Conversion

Add integer sign- and zero-extension and truncation to standard.

2019-09-21 16:14:56 -07:00

Dialect

Specalize f32->i8/u8 Quanitization with C++ native arithmetic to optimize performance.

2019-09-22 22:07:51 -07:00

EDSC

Add custom builder for AffineIfOp

2019-09-06 11:03:03 -07:00

Examples

LLVM dialect: prefix auxiliary operations with "mlir."

2019-09-03 09:10:56 -07:00

IR

Add integer sign- and zero-extension and truncation to standard.

2019-09-21 16:14:56 -07:00

lib

[ODS] Add support for FloatElementsAttr

2019-09-21 09:45:15 -07:00

mlir-cpu-runner

Add missing lowering to CFG in mlir-cpu-runner + related cleanup

2019-09-01 11:33:22 -07:00

mlir-cuda-runner

JitRunner: support entry functions returning void

2019-08-20 07:46:17 -07:00

mlir-tblgen

[ODS] Add support for FloatElementsAttr

2019-09-21 09:45:15 -07:00

Pass

Update the IRPrinter instrumentation to work on non function/module operations.

2019-09-14 21:56:38 -07:00

Quantizer

…

SDBM

Overhaul the SDBM expression kind hierarchy

2019-09-16 08:16:06 -07:00

Target

Add integer sign- and zero-extension and truncation to standard.

2019-09-21 16:14:56 -07:00

Transforms

Upgrade/fix/simplify store to load forwarding

2019-09-21 10:08:56 -07:00

Unit

…

APITest.h

…

CMakeLists.txt

Move Linalg and VectorOps dialects to the Dialect subdir - NFC

2019-08-19 17:11:38 -07:00

lit.cfg.py

…

lit.site.cfg.py.in

Move Linalg and VectorOps dialects to the Dialect subdir - NFC

2019-08-19 17:11:38 -07:00