Home
last modified time | relevance | path

Searched refs:fixedpoint (Results 1 – 3 of 3) sorted by relevance

/llvm-project-15.0.7/mlir/include/mlir/Dialect/Quant/
H A DUniformSupport.h112 APFloat fixedpoint = llvm::minimum(scaled, clampMax); in quantizeFloatToInt() local
113 fixedpoint = llvm::maximum(fixedpoint, clampMin); in quantizeFloatToInt()
116 fixedpoint.convertToInteger(result, roundMode, &lossy); in quantizeFloatToInt()
/llvm-project-15.0.7/mlir/test/Dialect/Quant/
H A Dconvert-const.mlir35 // Verifies i8 fixedpoint quantization on a splat tensor.
47 // Verifies i8 fixedpoint quantization on a splat tensor resulting in a negative storage value.
58 // Verifies i8 fixedpoint quantization on a dense tensor, sweeping values.
69 // Verifies i8 fixedpoint quantization on a sparse tensor, sweeping values.
83 // Verifies i8 fixedpoint quantization on a primitive const.
130 // Verifies i8 fixedpoint quantization on a dense tensor, sweeping values, and
H A Dparse-uniform.mlir141 // Per-axis scales and no zero points (fixedpoint)
150 // Per-axis scales and zero points (mixed affine and fixedpoint)