Introduction

This document describes how to develop LlamaIndex on Kunpeng servers, including source code build, installation, and development verification.

LlamaIndex is a data framework designed for large language model applications, providing data loading, indexing, retrieval, and orchestration capabilities. The homogeneous multi-core and high-concurrency capabilities of the Arm architecture of Kunpeng servers significantly accelerate intensive computing tasks such as document parsing, vector index construction, and retrieval.