Stars
Qwen3 is the large language model series developed by Qwen team, Alibaba Cloud.
Mind2Web-2 Benchmark: Evaluating Agentic Search with Agent-as-a-Judge
verl: Volcano Engine Reinforcement Learning for LLMs
DeepEP: an efficient expert-parallel communication library
KernelBench: Can LLMs Write GPU Kernels? - Benchmark with Torch -> CUDA problems
Official repository for FLAME-MoE: A Transparent End-to-End Research Platform for Mixture-of-Experts Language Models
LLM based autonomous agent that conducts deep local and web research on any topic and generates a long report with citations.
Official repository for "Craw4LLM: Efficient Web Crawling for LLM Pretraining"
🌐 WebThinker: Empowering Large Reasoning Models with Deep Research Capability
FastVideo is a unified framework for accelerated video generation.
The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery 🧑🔬
Attribute (or cite) statements generated by LLMs back to in-context information.
Fully open reproduction of DeepSeek-R1
Computer Vision tags on all 22 football film
Script to facilitate batch downloading of lecture videos from Panopto
Code for paper "Optima: Optimizing Effectiveness and Efficiency for LLM-Based Multi-Agent System"
HLLM: Enhancing Sequential Recommendations via Hierarchical Large Language Models for Item and User Modeling
Learning clinical-decision rules with interpretable models.
RWKV (pronounced RwaKuv) is an RNN with great LLM performance, which can also be directly trained like a GPT transformer (parallelizable). We are at RWKV-7 "Goose". So it's combining the best of RN…
The paper list of the review on LLMs in medicine - "Large Language Models Illuminate a Progressive Pathway to Artificial Healthcare Assistant: A Review".
An Open-source Framework for Data-centric, Self-evolving Autonomous Language Agents
Official repository for "Scaling Retrieval-Based Langauge Models with a Trillion-Token Datastore".
AI Logging for Interpretability and Explainability🔬
`dattri` is a PyTorch library for developing, benchmarking, and deploying efficient data attribution algorithms.
Minimalistic large language model 3D-parallelism training