Sign in with GitHub

← Back to Discover

ztxz16/fastllm

C++Apache-2.0active

Health

fastllm是后端无依赖的高性能大模型推理库。同时支持张量并行推理稠密模型和混合模式推理MOE模型，任意10G以上显卡即可推理满血DeepSeek。双路9004/9005服务器+单显卡部署DeepSeek满血满精度原版模型，单并发20tps；INT4量化模型单并发30tps，多并发可达60+。

Stars4.6k

Forks461

Open Issues287

Contributors461

Last Push3d ago

Health Breakdown

Activity

25

Community

13

Maintenance

13

Popularity

25

View on GitHub ↗Issues (287) ↗Pull Requests ↗Wiki ↗

Community

C++Apache 2.0

fastllm是后端无依赖的高性能大模型推理库。同时支持张量并行推理稠密模型和混合模式推理MOE模型，任意10G以上显卡即可推理满血DeepSeek。双路9004/9005服务器+单显卡部署DeepSeek满血满精度原版模型，单并发20tps；INT4量化模型单并发30tps，多并发可达60+。

active

★ 4.6k461 contributors3d ago

More C++ repos

MarlinFirmware/Marlin

Marlin is a firmware for RepRap 3D printers optimized for both 8 and 32 bit microcontrollers. Marlin supports all common platforms. Many commercial 3D printers come with Marlin installed. Check with your vendor if you need source code for your specific machine.

PX4/PX4-Autopilot

PX4 Autopilot Software

mavlink/qgroundcontrol

Cross-platform ground control station for drones (Android, iOS, Mac OS, Linux, Windows)