Fix QuantizedLinear for non-affine quantization modes (mxfp4)#384

davidkoski merged 1 commit intoml-explore:mainml-explore/mlx-swift:mainfrom

Nicolas-nwb:fix/quantized-linear-mxfp4Nicolas-nwb/mlx-swift:fix/quantized-linear-mxfp4Copy head branch name to clipboard