SGLang is a fast serving framework for large language models and vision language models.

GenRM-CoT: Data release for verification rationales

An Easy-to-use, Scalable and High-performance RLHF Framework (70B+ PPO Full Tuning & Iterative DPO & LoRA & RingAttention & RFT)

CLUEWSC2020: WSC Winograd模式挑战中文版，中文指代消解任务

InsTag: A Tool for Data Analysis in LLM Supervised Fine-tuning

CUDA Templates for Linear Algebra Subroutines

T2Ranking: A large-scale Chinese benchmark for passage ranking.

Google Research

DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.

A collection of notebooks/recipes showcasing some fun and effective ways of using Claude.

code for paper 《RankingGPT: Empowering Large Language Models in Text Ranking with Progressive Enhancement》

Lighteval is your all-in-one toolkit for evaluating LLMs across multiple backends

Minimalistic large language model 3D-parallelism training

Freeing data processing from scripting madness by providing a set of platform-agnostic customizable pipeline processing blocks.

Material for gpu-mode lectures

Efficient Triton Kernels for LLM Training

[ICML 2024] LESS: Selecting Influential Data for Targeted Instruction Tuning

Generate textbook-quality synthetic LLM pretraining data

DSIR large-scale data selection framework for language model training

General technology for enabling AI capabilities w/ LLMs and MLLMs

An open source trusted cloud native registry project that stores, signs, and scans content.

Cool Papers - Immersive Paper Discovery

Finetune Llama 3.3, DeepSeek-R1 & Reasoning LLMs 2x faster with 70% less memory

Build resilient language agents as graphs.

Easy, fast, and cheap pretrain,finetune, serving for everyone