Open-source projects and research initiatives incubated by LMSYS, building accessible large AI models, systems, and evaluation tools.
Systems
A fast serving engine for LLMs and VLMs.
An open and scalable platform for training, finetuning, serving, and evaluating LLM-based chatbots.
Train speculative decoding models effortlessly and port them smoothly to SGLang serving.
A system for serving thousands of concurrent LoRA adapters.
A framework for serving and evaluating LLM routers.
An exact, fast, parallel decoding algorithm without needing draft models or data stores.
Evaluation
A benchmark platform for LLMs that features anonymous, randomized battles in a crowdsourced manner.
An automatic pipeline converting live data to high quality benchmarks for evaluating chat assistants.
A set of challenging, multi-turn, and open-ended questions for evaluating chat assistants.
