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。

Making large AI models cheaper, faster and more accessible

An elegant PyTorch deep reinforcement learning library.

Tile primitives for speedy kernels

Ray is an AI compute engine. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.

verl: Volcano Engine Reinforcement Learning for LLMs

NanoGPT (124M) in 3 minutes

提取微信聊天记录，将其导出成HTML、Word、Excel文档永久保存，对聊天记录进行分析生成年度聊天报告，用聊天数据训练专属于个人的AI聊天助手

Community fork of PlayCover

Production-tested AI infrastructure tools for efficient AGI development and community-driven innovation

📚200+ Tensor/CUDA Cores Kernels, ⚡️flash-attn-mma, ⚡️hgemm with WMMA, MMA and CuTe (98%~100% TFLOPS of cuBLAS/FA2 🎉🎉).

A library for advanced large language model reasoning

解决Cursor在免费订阅期间出现以下提示的问题: You've reached your trial request limit. / Too many free trial accounts used on this machine. Please upgrade to pro. We have this limit in place to prevent abuse. Please l…

arXiv LaTeX Cleaner: Easily clean the LaTeX code of your paper to submit to arXiv

[ICLR 2025] DuoAttention: Efficient Long-Context LLM Inference with Retrieval and Streaming Heads

Repository for Label Words are Anchors: An Information Flow Perspective for Understanding In-Context Learning

Textbook on reinforcement learning from human feedback

Modeling, training, eval, and inference code for OLMo

Scalable RL solution for advanced reasoning of language models

BitBLAS is a library to support mixed-precision matrix multiplications, especially for quantized LLM deployment.

用于解决白嫖Cursor被封设备，适用windows、macos

A unified evaluation framework for large language models

A highly capable 2.4B lightweight LLM using only 1T pre-training data with all details.

Code for Paper: Training Software Engineering Agents and Verifiers with SWE-Gym

DeepSeek LLM: Let there be answers

lightweight, standalone C++ inference engine for Google's Gemma models.