Skip to main content

A lightweight, effective and easy-to-extend inference runtime

Project description

xlite logo

Xlite轻量化推理运行时

介绍

xlite (GVirt前端):轻量级Transformer模型运行时,支持多样性算力协同,当前支持在昇腾硬件上高效运行。 xlite公开了Transformer运行所需的模型构图以及算子,所有算子基于昇腾AscendC/CCE开发。

背景与动机

在大模型推理场景中,传统的单流串行执行模式存在以下问题:

  • 核间负载不均:不同AICORE之间的任务分配不均衡,部分核心闲置
  • 资源浪费多:计算资源和传输资源利用率低,存在显著浪费
  • 执行时间长:Host CPU下发算子开销大,造成严重的host bond问题

xlite通过以下技术手段解决上述问题:

  • 多流并行:充分利用卡内资源,将单流串行改为多流并行执行
  • 核间负载均衡:核间负载均衡,提升资源利用率
  • CPU NPU协同:C++侧完全消除Python的GC、线程等干扰,简化Host tiling计算,去除小块内存申请释放及拷贝,消除Host bond

性能效果显著:

在GLM-4.7双机推理场景(40K输入、1K输出、prefix cache命中率约90%)下:

  • TPOT时延降低17%~30%
  • 吞吐提升13%~41%

详细性能数据参考 PR #7935

软件架构

image

xlite已适配vllm_ascend,可通过xlite_graph_config配置快速使能xlite加速效果,使用方法参考官方指导文档

xlite支持模型见模型列表

快速开始

安装

# 安装vllm_ascend, 可参考https://github.com/vllm-project/vllm-ascend/blob/main/README.md
# 安装xlite
pip install xlite

离线推理示例

import os
from vllm import LLM

# xlite默认支持decode-only模式, 可通过设置 "full_mode": True 使能full模式
model = LLM(model="path/to/Qwen3-32B", tensor_parallel_size=8, additional_config={"xlite_graph_config": {"enabled": True, "full_mode": True}})
outputs = model.generate("Hello, how are you?")

在线服务示例

vllm serve path/to/Qwen3-32B --tensor-parallel-size 8 --additional-config='{"xlite_graph_config": {"enabled": true, "full_mode": true}}'

开发指南

编译构建、容器镜像、源码安装等开发相关内容,请参考 开发指南

环境变量

xlite使用的环境变量及其配置方法,请参考 环境变量

目录结构

.
├── csrc/    - 轻量化运行时的核心代码
├── xlite/   - python代码,包括tools等
├── doc/     - 相关文档介绍
├── docker/  - 容器镜像Dockerfile
└── tests/   - 测试用例

致谢

本项目的核心算子初始版本和思路来自于华为终端小艺AI Infra团队的贡献,相关优化实现可参考论文:

《XY-Serve: End-to-End Versatile Production Serving for Dynamic LLM Workloads》 [ASPLOS 2026]

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distributions

No source distribution files available for this release.See tutorial on generating distribution archives.

Built Distributions

If you're not sure about the file name format, learn more about wheel file names.

xlite-0.1.0rc11.dev210-cp312-cp312-manylinux2014_x86_64.whl (839.1 kB view details)

Uploaded CPython 3.12

xlite-0.1.0rc11.dev210-cp312-cp312-manylinux2014_aarch64.whl (787.1 kB view details)

Uploaded CPython 3.12

xlite-0.1.0rc11.dev210-cp311-cp311-manylinux2014_x86_64.whl (827.5 kB view details)

Uploaded CPython 3.11

xlite-0.1.0rc11.dev210-cp311-cp311-manylinux2014_aarch64.whl (786.0 kB view details)

Uploaded CPython 3.11

File details

Details for the file xlite-0.1.0rc11.dev210-cp312-cp312-manylinux2014_x86_64.whl.

File metadata

File hashes

Hashes for xlite-0.1.0rc11.dev210-cp312-cp312-manylinux2014_x86_64.whl
Algorithm Hash digest
SHA256 24dfc97b3b73368241bf16c271ec6c68411544725ff627adb279b536f6c53181
MD5 cc1dfa70d0153c6c62ca93a4aaffd939
BLAKE2b-256 33812dc782cd2ef1e4578a844f2bf2cde61b32c9dcc194ab7e6bd9a783f0ff35

See more details on using hashes here.

File details

Details for the file xlite-0.1.0rc11.dev210-cp312-cp312-manylinux2014_aarch64.whl.

File metadata

File hashes

Hashes for xlite-0.1.0rc11.dev210-cp312-cp312-manylinux2014_aarch64.whl
Algorithm Hash digest
SHA256 224e8b7abe767533631ef9d1cdf04708978e5a5ab9b963703c772e89d0bd9c94
MD5 dd28c5396d1a4652f22538fd4eeb91b2
BLAKE2b-256 44d26bc25f72dc2e3fda08a43785f490cbb06efe3f43602751a05efcd9147837

See more details on using hashes here.

File details

Details for the file xlite-0.1.0rc11.dev210-cp311-cp311-manylinux2014_x86_64.whl.

File metadata

File hashes

Hashes for xlite-0.1.0rc11.dev210-cp311-cp311-manylinux2014_x86_64.whl
Algorithm Hash digest
SHA256 b2e2f9794579e503804a07e8ba8aa4ffa0dcd4d100c39e0515aa21d90957f387
MD5 e321f036d131c50ca4dbb6d463455c0c
BLAKE2b-256 aeda606cbfaeb81cc14e9d4effa955d65e41be3e2a2be1d591c9e176870b3a21

See more details on using hashes here.

File details

Details for the file xlite-0.1.0rc11.dev210-cp311-cp311-manylinux2014_aarch64.whl.

File metadata

File hashes

Hashes for xlite-0.1.0rc11.dev210-cp311-cp311-manylinux2014_aarch64.whl
Algorithm Hash digest
SHA256 92d76a2a4cdc7e061563045313224cbe323a87b820d751dfcec2e0af24e7d98c
MD5 e8ac126d3df6626aa267215956d787e3
BLAKE2b-256 a81e43759f9769d7f1b0af398c6c15f8adf00c6cf6c4f48846673a2b131a560c

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page