Chansung Park

tunix. A Lightweight LLM Post-Training Library

CW. Python

diffusers-pipeline-autoresearch. Agentic AutoResearch scaffold for modular Diffusers experiments on JarvisLabs

tunix-accel. https://deep-diver.github.io/tunix-accel/

google-colab-cli. Python

dllm-jax. dLLM training implementation on pure jax/flax (w/o pytorch) for Google TPUs(v4/v5e/v6e). #TPUSprint #TRC

keras-diffusion-lab. Diffusion model research harness built on Keras 3 + JAX — unconditional DDPM, classifier-free guidance, and more

ReflexiCoder. [ACL'26] Official Code for "ReflexiCoder: Teaching Large Language Models to Self-Reflect on Generated Code and Self-Correct It via Reinforcement Learning"

WebGen-R1. [arXiv'26] Official Code for "WebGen-R1: Incentivizing Large Language Models to Generate Functional and Aesthetic Websites with Reinforcement Learning"

kinetic. Run ML workloads seamlessly on cloud TPUs and GPUs with a single Python decorator. No infrastructure management required.

NeuralComputer. 🖥 Neural Computers' Data Engine

MUSE. A library for Multilingual Unsupervised or Supervised word Embeddings

harness. A meta-skill that designs domain-specific agent teams, defines specialized agents, and generates the skills they use.

discord-daily-summarizer. Python

x-algorithm. Algorithm powering the For You feed on X

evalchemy-multipl-e-runner. Python

SurfMate. JavaScript

mini-sglang. A compact implementation of SGLang, designed to demystify the complexities of modern LLM serving systems.

a2ui. TypeScript

realtime-phone-agents-course. Build realtime AI voice agents using FastRTC for low-latency streaming, Superlinked for vector search, Twilio for live phone calls, and Runpod for scalable GPU deployment.

ai-fashion-moodboard. Generate unique fashion moodboards to find your next design inspiration.

arxiv2poster. Python

prime-rl. Agentic RL Training at Scale

Dream. Dream 7B, a large diffusion language model

trl. Train transformer language models with reinforcement learning.

dFactory. Easy and Efficient dLLM Fine-Tuning

verl. verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework

AlexNet. AlexNet model from ILSVRC 2012

nanochat. The best ChatGPT that $100 can buy.

tunix. A Lightweight LLM Post-Training Library

smol2operator. Python

Dream-Coder. Python

dLLM-RL. [ICLR 2026] Official code for TraceRL: Revolutionizing post-training for Diffusion LLMs, powering the SOTA TraDo series.

nano-banana-infinimap. TypeScript

awesome-nano-banana. Awesome curated collection of images and prompts generated by gemini-2.5-flash-image (aka Nano Banana) state-of-the-art image generation and editing model. Explore AI generated visuals created with Gemini, showcasing Google’s advanced image generation capabilities.

verifiers. Our library for RL environments + evals

gpt-oss-recipes. Collection of scripts and notebooks for OpenAI's latest GPT OSS models

hcx-vllm-plugin. vLLM plugin for HyperCLOVAX

ArchScale. Simple & Scalable Pretraining for Neural Architecture Research

EasyR1. EasyR1: An Efficient, Scalable, Multi-Modality RL Training Framework based on veRL

genai-processors. GenAI Processors is a lightweight Python library that enables efficient, parallel content processing.

Learning-JAX. Slide decks, coding exercises, and quick references for learning the JAX AI Stack

all-smi. Command-line utility for monitoring GPU hardware.

glass. Digital Mind Extension

auto-diffusers-config. Generate optimal code snippets to run diffursers models based on hardware setup

gemini-cli. An open-source AI agent that brings the power of Gemini directly into your terminal.

videoprism. Official repository for "VideoPrism: A Foundational Visual Encoder for Video Understanding" (ICML 2024)

auto-diffusers-docs. Use an LLM to generate reasonable code snippets in a hardware-aware manner for Diffusers.

Paper2Poster. [NeurIPS 2025] Open-source Multi-agent Poster Generation from Papers

aider. aider is AI pair programming in your terminal

containerization. Containerization is a Swift package for running Linux containers on macOS.

memvid. Memory layer for AI Agents. Replace complex RAG pipelines with a serverless, single-file memory layer. Give your agents instant retrieval and long-term memory.

genkit. Open-source framework for building AI-powered apps in JavaScript, Go, and Python, built and used in production by Google

gemini-fullstack-langgraph-quickstart. Get started with building Fullstack Agents using Gemini 2.5 and LangGraph

matformer. MatFormer repo

desktop. E2B Desktop Sandbox for LLMs. E2B Sandbox with desktop graphical environment that you can connect to any LLM for secure computer use.

langgraph-dynamic-mcp-agents. LangGraph ReAct Agents with an ability to use MCP Tools dynamically

pathways-utils. Package of Pathways-on-Cloud utilities

A2A. Agent2Agent (A2A) is an open protocol enabling communication and interoperability between opaque agentic applications.

Textoshop. 📝 An intelligent text editor with interactions inspired by drawing software

dynamo. A Datacenter Scale Distributed Inference Serving Framework

imm. Official implementation of Inductive Moment Matching

agent-chat-ui. 🦜💬 Web app for interacting with any LangGraph agent (PY & TS) via a chat interface.

openai-agents-python. A lightweight, powerful framework for multi-agent workflows

agentic-system. Jupyter Notebook

teddynote-parser-api-client. TeddyNote Parser API Client Library for Python

LeetCUDA. 📚LeetCUDA: Modern CUDA Learn Notes with PyTorch for Beginners🐑, 200+ CUDA Kernels, Tensor Cores, HGEMM, FA-2 MMA.🎉

omniserve. [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

AdaptSum. AdaptSum stands for Adaptive Summarization

kanana. Kanana: Compute-efficient Bilingual Language Models

fastrtc. The python library for real-time communication

scenic. Scenic: A Jax Library for Computer Vision Research and Beyond

DeepEP. DeepEP: an efficient expert-parallel communication library

kernels. Build compute kernels and load them from the Hub.

deep-research. An AI-powered research assistant that performs iterative, deep research on any topic by combining search engines, web scraping, and large language models. The goal of this repo is to provide the simplest implementation of a deep research agent - e.g. an agent that can refine its research direction overtime and deep dive into a topic.

mistral-common. Official inference library for pre-processing of Mistral models

open-r1. Fully open reproduction of DeepSeek-R1

DeepSeek-R1.

ComfyUI. The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface.

self-adaptive-llms. A Self-adaptation Framework🐙 that adapts LLMs for unseen tasks in real-time!

smolagents. 🤗 smolagents: a barebones library for agents that think in code.

minja. A minimalistic C++ Jinja templating engine for LLM chat templates

stable-diffusion-nano. Jupyter Notebook

picotron_tutorial. Python

DeepSeek-V3. Python

ModernBERT. Bringing BERT into modernity via both architecture changes and scaling

picotron. Minimalistic 4D-parallelism distributed training framework for education purpose

search-and-learn. Recipes to scale inference-time compute of open models

Repoformer. Repoformer: Selective Retrieval for Repository-Level Code Completion (ICML 2024)

python-genai. Google Gen AI Python SDK provides an interface for developers to integrate Google's generative models into their Python applications.

KaSA. [ICLR'25] Code for KaSA, an official implementation of "KaSA: Knowledge-Aware Singular-Value Adaptation of Large Language Models"

neurips2024. Read and Listen to NeurIPS 2024 Papers

CodeLLMSurvey. [TOSEM'25] The official GitHub page for the survey paper "A Survey on Large Language Models for Code Generation".

neurips2024-paper-reviewer. Python

advent-of-code-2024. Prolog

SK-WaveHill-2024. Jupyter Notebook

simple-image-recaptioning. Recaption large (Web)Datasets with vllm and save the artifacts.

hugo-theme-dream. 🌱 A Hugo theme named Dream.

logits-processor-zoo. A collection of LogitsProcessors to customize and enhance LLM behavior for specific tasks.

cs249r_book. Machine Learning Systems

paper-reviewer. Generate a comprehensive review from an arXiv paper, then turn it into a blog post. This project powers the website below for the HuggingFace's Daily Papers (https://huggingface.co/papers).

smollm. Everything about the SmolLM and SmolVLM family of models

notebookllama-as-module. Python

easy-pdf-rag. This is an example of implementing the RAG feature based on OpenAI Assistant V2 and Anthropic's PDF feature.

docling. Get your documents ready for gen AI

openinterpreter. A lightweight coding agent for open models like Deepseek, Kimi, and Qwen

ai-paper-reviewer. Auto generated blog posts of papers in the field of AI powered by "Paper Reviewer" project.

low-code-ai. Companion repository for "Low-Code AI" book

ogx. Open GenAI Stack

lectures. Material for gpu-mode lectures

read-paper-list. archive of read paper list

uv. An extremely fast Python package and project manager, written in Rust.

xLAM. xLAM: A Family of Large Action Models to Empower AI Agent Systems

RB-Modulation. Official code for "RB-Modulation: Training-Free Personalization of Diffusion Models using Stochastic Optimal Control"

writer-framework. No-code in the front, Python in the back. An open-source framework for creating data apps.

LitServe. A minimal Python framework for building custom AI inference servers with full control over logic, batching, and scaling.

CodeSolar. Making coding accessible to all with CodeSolar.

recall-mate. Python

fasthtml. The fastest way to create an HTML app

LLM101n. LLM101n: Let's build a Storyteller

GenAI-GeoGuesser. Generative AI version of the GeoGuesser game.

mesop. Rapidly build AI apps in Python

fine-tune-paligemma. Notebooks for fine tuning pali gemma

ftcourse. Jupyter Notebook

JLClient. Interacting with Jarvislabs.ai for creating GPU/CPU powered instances on top of A100, A6000, RTX 5000.

llama3-from-scratch. llama3 implementation one matrix multiplication at a time

llm-comparator. LLM Comparator is an interactive data visualization tool for evaluating and analyzing LLM responses side-by-side, developed by the PAIR team.

llamaduo-spinoff. Jupyter Notebook

genai-apis. Python

DigitalWindow. 착시 전자창문

penzai. A JAX research toolkit for building, editing, and visualizing neural networks.

llama3. The official Meta Llama 3 GitHub site

sonar. Large-scale LLM inference engine

promptengineer.

hllama. hllama is a library which aims to provide a set of utility tools for large language models.

JetStream. JetStream is a throughput and memory optimized engine for LLM inference on XLA devices, starting with TPUs (and GPUs in future -- PRs welcome).

lightning-whisper-mlx. An extremely fast implementation of whisper optimized for Apple Silicon using MLX.

unsloth. Unsloth Studio is a web UI for training and running open models like Gemma 4, Qwen3.6, DeepSeek, gpt-oss locally.

axolotl. Go ahead and axolotl questions

optimum-quanto. A pytorch quantization backend for optimum

AIOS. AIOS: AI Agent Operating System

moondream. tiny vision language model

torchtune. PyTorch native post-training library

jaxtyping. Type annotations and runtime checking for shape and dtype of JAX/NumPy/PyTorch/etc. arrays. https://docs.kidger.site/jaxtyping/

grok-1. Grok open release

markdown-badges-generator. Generate badges for your READMEs - inspired by ileriayo/markdown-badges

awesome-osml-for-devs. List of resources, libraries and more for developers who would like to build with open-source machine learning off-the-shelf

paperqa-ui. Python

fsdp_qlora. Training LLMs with QLoRA + FSDP

Vid2Persona. This project breathes life into video characters by using AI to describe their personality and then chat with you as them.

llamaduo. [ACL'25] Official Code for LlamaDuo: LLMOps Pipeline for Seamless Migration from Service LLMs to Small-Scale Local LLMs

GPTFast. Accelerate your Hugging Face Transformers 7.6-9x. Native to Hugging Face and PyTorch.

gemma.cpp. lightweight, standalone C++ inference engine for Google's Gemma models.

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

gemma. Gemma open-weight LLM library, from Google DeepMind

jepa. PyTorch code and models for V-JEPA self-supervised learning from video.

janus. generate synthetic data for LLM fine-tuning in arbitrary situations within systematic way

auto-paper-analysis. Jupyter Notebook

llama-keras. Jupyter Notebook

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

honeybee. Official implementation of project Honeybee (CVPR 2024)

PhotoMaker. PhotoMaker [CVPR 2024]

hf-daily-paper-newsletter. Newsletter bot for 🤗 Daily Papers

AMIE-pytorch. Implementation of the general framework for AMIE, from the paper "Towards Conversational Diagnostic AI", out of Google Deepmind

k3im. (K3IM) Keras 3 Image Models

Awesome-LLM-Inference. 📚A curated list of Awesome LLM/VLM Inference Papers with Codes: Flash-Attention, Paged-Attention, WINT8/4, Parallelism, etc.🎉

Yi. A series of large language models trained from scratch by developers @01-ai

keras_cv_attention_models. Keras beit,caformer,CMT,CoAtNet,convnext,davit,dino,efficientdet,edgenext,efficientformer,efficientnet,eva,fasternet,fastervit,fastvit,flexivit,gcvit,ghostnet,gpvit,hornet,hiera,iformer,inceptionnext,lcnet,levit,maxvit,mobilevit,moganet,nat,nfnets,pvt,swin,tinynet,tinyvit,uniformer,volo,vanillanet,yolor,yolov7,yolov8,yolox,gpt2,llama2, alias kecam

Awesome-LLM. Awesome-LLM: a curated list of Large Language Model

Zero2Story. Zero2Story

low-code-ai. Companion repository for "Low-Code AI" book

NimStoryFont. An application for storing your notes for your tabletop RPG campaigns!

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

llama-cookbook. Welcome to the Llama Cookbook! This is your go to guide for Building with Llama: Getting started with Inference, Fine-Tuning, RAG. We also show you how to solve end to end problems using Llama model family and using them on various provider services

LLM-As-Chatbot. LLM as a Chatbot Service

keras-core. A multi-backend implementation of the Keras API, with support for TensorFlow, JAX, and PyTorch.

LLM-Training-Puzzles. What would you do with 1000 H100s...

alpaca-roleplay-discordbot. A discord bot that roleplays!

DragDiffusion. Implementation of DragDiffusion: Harnessing Diffusion Models for Interactive Point-based Image Editing

keras-io. Keras documentation, hosted live at keras.io

sgpt. SGPT: GPT Sentence Embeddings for Semantic Search

gpt2-ft-pipeline. GPT2 fine-tuning pipeline with KerasNLP, TensorFlow, and TensorFlow Extended

chatui. ChatGPT directly in your terminal using Textual

song-describer. Song Describer is a data collection platform for annotating music with textual descriptions.

OpenChat. LLMs custom-chatbots console ⚡

KULLM. ☁️ 구름(KULLM): 고려대학교에서 개발한, 한국어에 특화된 LLM

LLM-Pref-Mark-UI. Python

LLMs-Colab. Python

instruction-tuned-sd. Code for instruction-tuning Stable Diffusion.

gradio-chat. HuggingChat like UI in Gradio

runtime. A performant and modular runtime for TensorFlow

bitsandbytes. Accessible large language models via k-bit quantization for PyTorch.

LLM-Pool. Python

vicuna-lora. Instruct-tune LLaMA on consumer hardware

flutter-tflite. Dart

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

IF. Python

flan-ul2-alpaca. Python

whisper-jax. JAX implementation of OpenAI's Whisper model for up to 70x speed-up on TPU.

picaisso. 🎨 Imagine what Picasso could have done with AI. Self-host your StableDiffusion API.