Skip to main content

quantize

Function quantize 

Source
pub(super) fn quantize(bytes: &[u8]) -> Result<Group>
Expand description

Use MLX’s affine endpoint selection and ties-to-even rounding. Quantize with float parameters, then store them as BF16, as MLX does for BF16 input. Reference: mlx/backend/cpu/quantized.cpp, quantize<T, U>.