Anonymous

FlashQLA. high-performance linear attention kernel library built on TileLang

TileKernels. A kernel library written in tilelang

ml-intern. 🤗 ml-intern: an open-source ML engineer that reads papers, trains models, and ships ML models

CubeSandbox. Instant, Concurrent, Secure & Lightweight Sandbox for AI Agents.

FlashKDA. FlashKDA: high-performance Kimi Delta Attention kernels

Awesome-LLM-Scientific-Discovery. [EMNLP2025] From Automation to Autonomy: A Survey on Large Language Models in Scientific Discovery

pinchtab. High-performance browser automation bridge and multi-instance orchestrator with advanced stealth injection and real-time dashboard.

SteptronOss. A lightweight, AI-native training framework for large language models. Designed for fast iteration, reproducible experiments, and modular configuration across SFT, RLVR, and evaluation workflows.

NextStep-1. [🚀 ICLR 2026 Oral] NextStep-1: SOTA Autogressive Image Generation with Continuous Tokens. A research project developed by the StepFun’s Multimodal Intelligence team.

slime. slime is an LLM post-training framework for RL Scaling.

DistRL-open. Python

vibetensor. Our first fully AI generated deep learning system

cosmos-reason2. Cosmos-Reason2 models understand the physical common sense and generate appropriate embodied decisions in natural language through long chain-of-thought reasoning processes.

cosmos-rl. Cosmos-RL is a flexible and scalable Reinforcement Learning framework specialized for Physical AI applications.

SimpleVLA-RL. [ICLR 2026] SimpleVLA-RL: Scaling VLA Training via Reinforcement Learning

openpi. Python

pipelining-sft. Simple and efficient DeepSeek V3 SFT using pipeline parallel and expert parallel, with both FP8 and BF16 trainings

action_piece. Python

linear-attention-and-beyond-slides. TeX

gpt-oss. gpt-oss-120b and gpt-oss-20b are two open-weight language models by OpenAI

alphagenome. This API provides programmatic access to the AlphaGenome model developed by Google DeepMind.

Liger-Kernel. Efficient Triton Kernels for LLM Training

ByteCheckpoint. ByteCheckpoint: An Unified Checkpointing Library for LFMs

Triton-distributed. Distributed Compiler based on Triton for Parallel Systems

DeepResearch. Tongyi Deep Research, the Leading Open-source Deep Research Agent

Megakernels. Kernels, of the mega variety :)

reverse-engineering-gemma-3n. Reverse Engineering Gemma 3n: Google's New Edge-Optimized Language Model

Spurious_Rewards. Python

ChinaTextbook. 所有小初高、大学PDF教材。

NeMo-Aligner. Scalable toolkit for efficient model alignment

transfusion-pytorch. Pytorch implementation of Transfusion, "Predict the Next Token and Diffuse Images with One Multi-Modal Model", from MetaAI

rllm. Democratizing Reinforcement Learning for LLMs

AReaL. The RL Bridge for LLM-based Agent Applications. Made Simple & Flexible.

understand-r1-zero. Understanding R1-Zero-Like Training: A Critical Perspective

DAPO. An Open-source RL System from ByteDance Seed and Tsinghua AIR

torchtitan. A PyTorch native platform for training generative AI models

open-r1. Fully open reproduction of DeepSeek-R1

flux. A fast communication-overlapping library for tensor/expert parallelism on GPUs.

OLMo-core. PyTorch building blocks for the OLMo ecosystem

smollm. Everything about the SmolLM and SmolVLM family of models

3FS. A high-performance distributed file system designed to address the challenges of AI training and inference workloads.

DualPipe. A bidirectional pipeline parallelism algorithm for computation-communication overlap in DeepSeek V3/R1 training.

Awesome-System2-Reasoning-LLM. Latest Advances on System-2 Reasoning

SpargeAttn. [ICML2025] SpargeAttention: A training-free sparse attention that accelerates any model inference.

DeepEP. DeepEP: an efficient expert-parallel communication library

vattention. Dynamic Memory Management for Serving LLMs without PagedAttention

FlashMLA. FlashMLA: Efficient Multi-head Latent Attention Kernels

Open-Reasoner-Zero. Official Repo for Open-Reasoner-Zero

RAGEN. RAGEN leverages reinforcement learning to train LLM reasoning agents in interactive, stochastic environments.

RecFlow. Python

BladeDISC. BladeDISC is an end-to-end DynamIc Shape Compiler project for machine learning workloads.

PQCache. [SIGMOD 2025] PQCache: Product Quantization-based KVCache for Long Context LLM Inference

Awesome-Multimodal-Large-Language-Models. :sparkles::sparkles:Latest Advances on Multimodal Large Language Models

trl. Train transformer language models with reinforcement learning.

Slow_Thinking_with_LLMs. A series of technical report on Slow Thinking with LLM

LRURec. [WSDM 2024] Official PyTorch Implementation of Linear Recurrent Units for Sequential Recommendation (LRURec)

DreamRec. Python

moe_attention. Official repository for the paper "SwitchHead: Accelerating Transformers with Mixture-of-Experts Attention"

MoH. MoH: Multi-Head Attention as Mixture-of-Head Attention

olmes. Reproducible, flexible LLM evaluations

open-instruct. AllenAI's post-training codebase

LLaMA-O1. Large Reasoning Models

nanoGPT-mup. The simplest, fastest repository for training/finetuning medium-sized GPTs.

mup. maximal update parametrization (µP)

DeepSeek-Coder. DeepSeek Coder: Let the Code Write Itself

open_llama. OpenLLaMA, a permissively licensed open source reproduction of Meta AI’s LLaMA 7B trained on the RedPajama dataset

OLMo. Modeling, training, eval, and inference code for OLMo

openr. OpenR: An Open Source Framework for Advanced Reasoning with Large Language Models

O1-Journey. O1 Replication Journey

baselines. OpenAI Baselines: high-quality implementations of reinforcement learning algorithms

quiet-star. Code for Quiet-STaR

triton. Development repository for the Triton language and compiler

alphageometry. Python

xformers. Hackable and optimized Transformers building blocks, supporting a composable construction.

grade-school-math. Python

DCFormer. Python

math. The MATH Dataset (NeurIPS 2021)

gpu.cpp. A lightweight library for portable low-level GPU computation using WebGPU.

TransformerEngine. A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit and 4-bit floating point (FP8 and FP4) precision on Hopper, Ada and Blackwell GPUs, to provide better performance with lower memory utilization in both training and inference.

dlrover. DLRover: An Automatic Distributed Deep Learning System

OpenMoE. A family of open-sourced Mixture-of-Experts (MoE) Large Language Models

graphrag. A modular graph-based Retrieval-Augmented Generation (RAG) system

megablocks. Python

lingvo. Lingvo

rq-vae-transformer. The official implementation of Autoregressive Image Generation using Residual Quantization (CVPR '22)

Index-1.9B. A lightweight multilingual LLM

k2-train. Python

MOSS-RLHF. Secrets of RLHF in Large Language Models Part I: PPO

nni. An open source AutoML toolkit for automate machine learning lifecycle, including feature engineering, neural architecture search, model compression and hyper-parameter tuning.

generative-recommenders. Repository hosting code for "Actions Speak Louder than Words: Trillion-Parameter Sequential Transducers for Generative Recommendations" (https://arxiv.org/abs/2402.17152).

llm.c. LLM training in simple, raw C/CUDA

ragas. Supercharge Your LLM Application Evaluations 🚀

LLaVA. [NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.

semantic-kernel. Integrate cutting-edge LLM technology quickly and easily into your apps

FlagEmbedding. Retrieval and Retrieval-augmented LLMs

text-embeddings-inference. A blazing fast inference solution for text embeddings models

promptflow. Build high-quality LLM apps - from prototyping, testing to production deployment and monitoring.

PokemonRedExperiments. Playing Pokemon Red with Reinforcement Learning

streaming-llm. [ICLR 2024] Efficient Streaming Language Models with Attention Sinks

DocsGPT. Private AI platform for agents, assistants and enterprise search. Built-in Agent Builder, Deep research, Document analysis, Multi-model support, and API connectivity for agents.

ToRA. ToRA is a series of Tool-integrated Reasoning LLM Agents designed to solve challenging mathematical reasoning problems by interacting with tools [ICLR'24].

LanguageAgentTreeSearch. [ICML 2024] Official repository for "Language Agent Tree Search Unifies Reasoning Acting and Planning in Language Models"

dspy. DSPy: The framework for programming—not prompting—language models

exllama. A more memory-efficient rewrite of the HF transformers implementation of Llama for use with quantized weights.

gorilla. Gorilla: Training and Evaluating LLMs for Function Calls (Tool Calls)

LightLLM. LightLLM is a Python-based LLM (Large Language Model) inference and serving framework, notable for its lightweight design, easy scalability, and high-speed performance.

quivr. Opiniated RAG for integrating GenAI in your apps 🧠 Focus on your product rather than the RAG. Easy integration in existing products with customisation! Any LLM: GPT4, Groq, Llama. Any Vectorstore: PGVector, Faiss. Any Files. Anyway you want.

Platypus. Code for fine-tuning Platypus fam LLMs using LoRA

AgentBench. A Comprehensive Benchmark to Evaluate LLMs as Agents (ICLR'24)

llama2.c. Inference Llama 2 in one file of pure C

AutoGPT. AutoGPT is the vision of accessible AI for everyone, to use and to build on. Our mission is to provide the tools, so that you can focus on what matters.

InternLM. Official release of InternLM series (InternLM, InternLM2, InternLM2.5, InternLM3).

illa-builder. Low-code platform allows you to build business apps, enables you to quickly create internal tools such as dashboard, crud app, admin panel, crm, cms, etc. Supports PostgreSQL, MySQL, Supabase, GraphQL, MongoDB, MSSQL, Rest API, Hugging Face, Redis, etc. Automate workflows with schedule or webhook. Open source Retool.

Grounding_LLMs_with_online_RL. We perform functional grounding of LLMs' knowledge in BabyAI-Text

ReAct. [ICLR 2023] ReAct: Synergizing Reasoning and Acting in Language Models

lmdeploy. LMDeploy is a toolkit for compressing, deploying, and serving LLMs.

vllm. A high-throughput and memory-efficient inference and serving engine for LLMs

MetaGPT. 🌟 The Multi-Agent Framework: First AI Software Company, Towards Natural Language Programming

ChatLaw. ChatLaw：A Powerful LLM Tailored for Chinese Legal. 中文法律大模型

Firefly. Firefly: 大模型训练工具，支持训练Qwen2.5、Qwen2、Yi1.5、Phi-3、Llama3、Gemma、MiniCPM、Yi、Deepseek、Orion、Xverse、Mixtral-8x7B、Zephyr、Mistral、Baichuan2、Llma2、Llama、Qwen、Baichuan、ChatGLM2、InternLM、Ziya2、Vicuna、Bloom等大模型

stable-diffusion. A latent text-to-image diffusion model

LLM-Blender. [ACL2023] We introduce LLM-Blender, an innovative ensembling framework to attain consistently superior performance by leveraging the diverse strengths of multiple open-source LLMs. LLM-Blender cut the weaknesses through ranking and integrate the strengths through fusing generation to enhance the capability of LLMs.

FasterViT. [ICLR 2024] Official PyTorch implementation of FasterViT: Fast Vision Transformers with Hierarchical Attention

InternLM-techreport.

Youku-mPLUG. Youku-mPLUG: A 10 Million Large-scale Chinese Video-Language Pre-training Dataset and Benchmarks

YuLan-Chat. YuLan: An Open-Source Large Language Model

TigerBot. TigerBot: A multi-language multi-task LLM

awesome-chinese-legal-resources. 📝 An Awesome Collection of Chinese Legal Dataset and Relevant Resources. 致力于收集全面的中文法律数据源

DragGAN. Official Code for DragGAN (SIGGRAPH 2023)

cc_net. Tools to download and cleanup Common Crawl data

lm-evaluation-harness. A framework for few-shot evaluation of language models.

criu. Checkpoint/Restore tool

CoT-Collection. [EMNLP 2023] The CoT Collection: Improving Zero-shot and Few-shot Learning of Language Models via Chain-of-Thought Fine-Tuning

lit-llama. Implementation of the LLaMA language model based on nanoGPT. Supports flash attention, Int8 and GPTQ 4bit quantization, LoRA and LLaMA-Adapter fine-tuning, pre-training. Apache 2.0-licensed.

llm-foundry. LLM training code for Databricks foundation models

PandaGPT. [TLLM'23] PandaGPT: One Model To Instruction-Follow Them All

pile_dedupe. Pile Deduplication Code

MaskDINO. [CVPR 2023] Official implementation of the paper "Mask DINO: Towards A Unified Transformer-based Framework for Object Detection and Segmentation"

List-of-Dirty-Naughty-Obscene-and-Otherwise-Bad-Words. List of Dirty, Naughty, Obscene, and Otherwise Bad Words

bigscience. Central place for the engineering/scaling WG: documentation, SLURM scripts and logs, compute environment and data.

qlora. QLoRA: Efficient Finetuning of Quantized LLMs

tree-of-thoughts. Plug in and Play Implementation of Tree of Thoughts: Deliberate Problem Solving with Large Language Models that Elevates Model Reasoning by atleast 70%

LLMSurvey. The official GitHub page for the survey paper "A Survey of Large Language Models".

Linly. Chinese-LLaMA 1&2、Chinese-Falcon 基础模型；ChatFlow中文对话模型；中文OpenLLaMA模型；NLP预训练/指令微调数据集

RL4LMs. A modular RL library to fine-tune language models to human preferences

WizardLM. LLMs build upon Evol Insturct: WizardLM, WizardCoder, WizardMath

unlimiformer. Public repo for the NeurIPS 2023 paper "Unlimiformer: Long-Range Transformers with Unlimited Length Input"

chain-of-thought-hub. Benchmarking large language models' complex reasoning ability with chain-of-thought prompting

alpa. Training and serving large-scale neural networks with auto parallelization.

InstructionZoo.

LLMZoo. ⚡LLM Zoo is a project that provides data, models, and evaluation benchmark for large language models.⚡

chameleon-llm. Codes for "Chameleon: Plug-and-Play Compositional Reasoning with Large Language Models".

MOSS. An open-source tool-augmented conversational language model from Fudan University

StableLM. StableLM: Stability AI Language Models

MiniGPT-4. Open-sourced codes for MiniGPT-4 and MiniGPT-v2 (https://minigpt-4.github.io, https://minigpt-v2.github.io/)

hh-rlhf. Human preference data for "Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback"

randomfun. Notebooks and various random fun

dolly. Databricks’ Dolly, a large language model trained on the Databricks Machine Learning Platform

TencentPretrain. Tencent Pre-training framework in PyTorch & Pre-trained Model Zoo

helm. Holistic Evaluation of Language Models (HELM) is an open source Python framework created by the Center for Research on Foundation Models (CRFM) at Stanford for holistic, reproducible and transparent evaluation of foundation models, including large language models (LLMs) and multimodal models.

Llama-X. Open Academic Research on Improving LLaMA to SOTA LLM

MarkTool. DoTAT 是一款基于web、面向领域的通用文本标注工具，支持大规模实体标注、关系标注、事件标注、文本分类、基于字典匹配和正则匹配的自动标注以及用于实现归一化的标准名标注，同时也支持迭代标注、嵌套实体标注和嵌套事件标注。标注规范可自定义且同类型任务中可“一次创建多次复用”。通过分级实体集合扩大了实体类型的规模，并设计了全新高效的标注方式，提升了用户体验和标注效率。此外，本工具增加了审核环节，可对多人的标注结果进行一致性检验、自动合并和手动调整，提高了标注结果的准确率。

LMFlow. An Extensible Toolkit for Finetuning and Inference of Large Foundation Models. Large Models for All.

JARVIS. JARVIS, a system to connect LLMs with ML community. Paper: https://arxiv.org/pdf/2303.17580.pdf

FasterTransformer. Transformer related optimization, including BERT, GPT

the-algorithm. Source code for the X Recommendation Algorithm

FastChat. An open platform for training, serving, and evaluating large language models. Release repo for Vicuna and Chatbot Arena.

BELLE. BELLE: Be Everyone's Large Language model Engine（开源中文对话大模型）

LoRA. Code for loralib, an implementation of "LoRA: Low-Rank Adaptation of Large Language Models"

Luotuo-Chinese-LLM. 骆驼(Luotuo): Open Sourced Chinese Language Models. Developed by 陈启源 @ 华中师范大学 & 李鲁鲁 @ 商汤科技 & 冷子昂 @ 商汤科技

pytorch-vit. An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale

ChatDoctor. Python

Chinese-LLaMA-Alpaca. 中文LLaMA&Alpaca大语言模型+本地CPU/GPU训练部署 (Chinese LLaMA & Alpaca LLMs)

chatgpt-retrieval-plugin. The ChatGPT Retrieval Plugin lets you easily find personal or work documents by asking questions in natural language.

alpaca-lora. Instruct-tune LLaMA on consumer hardware

document.ai. 基于向量数据库与GPT3.5的通用本地知识库方案(A universal local knowledge base solution based on vector database and GPT3.5)

evals. Evals is a framework for evaluating LLMs and LLM systems, and an open-source registry of benchmarks.

gpt-neox. An implementation of model parallel autoregressive transformers on GPUs, based on the Megatron and DeepSpeed libraries

Fengshenbang-LM. Fengshenbang-LM(封神榜大模型)是IDEA研究院认知计算与自然语言研究中心主导的大模型开源体系，成为中文AIGC和认知智能的基础设施。

ChatGLM-6B. ChatGLM-6B: An Open Bilingual Dialogue Language Model | 开源双语对话语言模型

stanford_alpaca. Code and documentation to train Stanford's Alpaca models, and generate the data.

taichi. Productive, portable, and performant GPU programming in Python.

TaskMatrix. Python

optimate. A collection of libraries to optimise AI model performances

GLM-130B. GLM-130B: An Open Bilingual Pre-Trained Model (ICLR 2023)

CVPR2026-Papers-with-Code. CVPR 2026 论文和开源项目合集

OpenDelta. A plug-and-play library for parameter-efficient-tuning (Delta Tuning)

langchain. The agent engineering platform.

llama_index. LlamaIndex is the leading document agent and OCR platform

paper-qa. High accuracy RAG for answering questions from scientific documents with citations

open_clip. An open source implementation of CLIP.

trlx. A repo for distributed training of language models with Reinforcement Learning via Human Feedback (RLHF)

PaLM-rlhf-pytorch. Implementation of RLHF (Reinforcement Learning with Human Feedback) on top of the PaLM architecture. Basically ChatGPT but with PaLM

pai. Resource scheduling and cluster management for AI

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

Megatron-LM. Ongoing research training transformer models at scale

gluten. Gluten is a middle layer responsible for offloading JVM-based SQL engines' execution to native engines.

latent-diffusion. High-Resolution Image Synthesis with Latent Diffusion Models

miniF2F. An updated version of miniF2F with lots of fixes and informal statements / solutions.

ControlNet. Let us control diffusion models!