Stars
Simple, scalable AI model deployment on GPU clusters
Gin is a high-performance HTTP web framework written in Go. It provides a Martini-like API but with significantly better performance—up to 40 times faster—thanks to httprouter. Gin is designed for …
Production-ready platform for agentic workflow development.
Facilitates running Wasm / WASI workloads managed by containerd
Serverless LLM Serving for Everyone.
User-friendly AI Interface (Supports Ollama, OpenAI API, ...)
TiRex: Zero-Shot Forecasting Across Long and Short Horizons with Enhanced In-Context Learning
[EMNLP2025] "LightRAG: Simple and Fast Retrieval-Augmented Generation"
The official Rust implementation of Conflux protocol. https://doc.confluxnetwork.org
Enjoy the magic of Diffusion models!
Chronos: Pretrained Models for Time Series Forecasting
AIInfra（AI 基础设施）指AI系统从底层芯片等硬件，到上层软件栈支持AI大模型训练和推理。
real time face swap and one-click video deepfake with only a single image
Kronos: A Foundation Model for the Language of Financial Markets
SeBS: serverless benchmarking suite for automatic performance analysis of FaaS platforms.
Official reinforcement learning environment for demand response and load shaping
Unified Training of Universal Time Series Forecasting Transformers
Official code for "F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching"
A high-performance ML model serving framework, offers dynamic batching and CPU/GPU pipelines to fully exploit your compute machine
A clean, elegant, beautiful and powerful admin template, based on Vue3, Vite7, TypeScript, Pinia, NaiveUI and UnoCSS. 一个清新优雅、高颜值且功能强大的后台管理模板，基于最新的前端技术栈，包括 Vue3, Vite7, TypeScript, Pinia, NaiveUI 和 …
An open-source cross-platform alternative to AirDrop
Framework that integrates the serverless benchmark suite vSwarm with gem5, the state-of-the-art research platform for system-and microarchitecture.
A framework for few-shot evaluation of language models.
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+。
OpenCompass is an LLM evaluation platform, supporting a wide range of models (Llama3, Mistral, InternLM2,GPT-4,LLaMa2, Qwen,GLM, Claude, etc) over 100+ datasets.