Function Description

AscendIndexIVFast contains the dimension reduction algorithm IVFSQ for three-level retrieval. Two parameters need to be passed to specify the dimension information before and after dimension reduction. The reduced dimension must be able to exactly divide the previous dimension.

Train an AI clustering model on the GPU.

Generate the operators required for the three-level retrieval based on the IVFSQT operator generation method.

In addition, this class also provides the fuzzy clustering function. Use the threshold parameter to control the base library redundancy before vectors enter a bucket. You are advised to set batch size to 65536 for search.