Stars
Miles is an enterprise-facing reinforcement learning framework for LLM and VLM post-training, forked from and co-evolving with slime.
[ICLR 2026] Information Gain-based Policy Optimization: A Simple and Effective Approach for Multi-Turn Search Agents
[NeurIPS 2025] Official Implementation of ViSpec: Accelerating Vision-Language Models with Vision-Aware Speculative Decoding.
how to optimize some algorithm in cuda.
Get started with building Fullstack Agents using Gemini 2.5 and LangGraph
ZeroSearch: Incentivize the Search Capability of LLMs without Searching
SGLang is a high-performance serving framework for large language models and multimodal models.
[MLSys'25] QServe: W4A8KV4 Quantization and System Co-design for Efficient LLM Serving; [MLSys'25] LServe: Efficient Long-sequence LLM Serving with Unified Sparse Attention
A concise but complete full-attention transformer with a set of promising experimental features from various papers
Efficient Triton Kernels for LLM Training
《李宏毅深度学习教程》（李宏毅老师推荐👍，苹果书🍎），PDF下载地址：https://github.com/datawhalechina/leedl-tutorial/releases
A pytorch quantization backend for optimum
Ongoing research training transformer models at scale
什么？你敢放心的把后背交给 AI? 我赌你不敢，那就来学学 AI 时代最酷、最安全、最快的语言吧。本书拥有全面且深入的讲解、生动贴切的示例、德芙般丝滑的内容，这可能是目前最用心的 Rust 中文学习教程 / Book
Making large AI models cheaper, faster and more accessible
中文nlp解决方案(大模型、数据、模型、训练、推理)
GLM-130B: An Open Bilingual Pre-Trained Model (ICLR 2023)
Transfer learning / domain adaptation / domain generalization / multi-task learning etc. Papers, codes, datasets, applications, tutorials.-迁移学习
PostgreSQL manual Chinese translation by China PostgreSQL Users Group
Code for "Gradient Surgery for Multi-Task Learning"
key Deep Learning engineering tricks in recsys
BlackHole is a modern macOS audio loopback driver that allows applications to pass audio to other applications with zero additional latency.