to_affine_quantized_fpx¶ torchao.dtypes.to_affine_quantized_fpx(input_float: Tensor, _layout: Layout)¶