Teïlo M

OScaR-KV-Quant. 🏆 OScaR: The Occam's Razor for Extreme KV Cache Quantization in LLMs and Beyond — redefining the accuracy-efficiency Pareto front for X-LLMs KV quantization.

reap. REAP: Router-weighted Expert Activation Pruning for SMoE compression

code. Noumena's internal version of your favorite coding tui but improved and optimized for our inference stack

RapidRAW. A beautiful, non-destructive, and GPU-accelerated RAW image editor built with performance in mind.

PawBench. A benchmark for evaluating LLM × harness performance.

lathe. Generate hands-on, multi-part technical tutorials on demand, with LLM skills tuned to make content approachable. Then you work through them yourself, by hand ✋

tracer. TRACER: replace 90%+ of your LLM classification calls with a traditional ML model. Formal parity guarantees. Self-improving.

optimum-executorch. 🤗 Optimum ExecuTorch

executorch. On-device AI across mobile, embedded and edge for PyTorch

causalab. Python

OpenShell. OpenShell is the safe, private runtime for autonomous AI agents.

rlix. Run more RL experiments. Wait less for GPUs.

gym-anything. Gym-Anything: Turn any Software into an Agent Environment

ordeal. Automated chaos testing for Python — fault injection, property assertions, and stateful exploration

immorteilo. Python

deltavar. https://arxiv.org/pdf/2502.14698

edgerunner. kelly criterion web apps

homebrew-mime. Homebrew tap for Mime - a distraction-free note-taking app

mime. journey taker

kromcanon. Steering vectors in non-standard transformer architectures (Canon + KromHC)

xyz. JavaScript

vauban. MLX-native toolkit for understanding and reshaping how language models behave on Apple Silicon

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

monty. A minimal, secure Python interpreter written in Rust for use by AI

black-box-query-bench. Python

enzu. Budgeted LLM runs with hard caps + typed outcomes + async job mode. Python/CLI/HTTP. OpenAI-compatible.

amla-sandbox. Python

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

PhysicsLM4. Physics of Language Models: Part 4.2, Canon Layers at Scale where Synthetic Pretraining Resonates in Reality

gepa. Optimize prompts, code, and more with AI-powered Reflective Text Evolution

thunnini. Experimentation library for comparing fine-tuners for neural sequential predictors.

EmberJson. A user friendly json library written in pure Mojo

nabla. Nabla: High-Performance Scientific Computing

tunix. A Lightweight LLM Post-Training Library

tinyworlds. A minimal implementation of DeepMind's Genie world model

tinker-cookbook. Post-training with Tinker

tinker. Training API and CLI

catgrad. a categorical deep learning compiler

LaTeX-graphics-with-TikZ. LaTeX graphics with TikZ, by Packt Publishing

RelationalThinking-Book. TeX

open-game-engine. Haskell implementation of open games

RelationalThinking-code. This repository contains all the code presented in the online book "Relational Thinking - from Abstractions to Applications".

paperproof. Lean theorem proving interface which feels like pen-and-paper proofs.

textpolicy. Reinforcement learning for text generation on MLX (Apple Silicon)

embedding-atlas. Embedding Atlas is a tool that provides interactive visualizations for large embeddings. It allows you to visualize, cross-filter, and search embeddings and metadata.

exo. Run frontier AI locally.

agent-lightning. The absolute trainer to light up AI agents.

Minigrid. Simple and easily configurable grid world environments for reinforcement learning

openbench. Provider-agnostic, open-source evaluation infrastructure for language models

gem. A Gym for Agentic LLMs

dreamerv3. Mastering Diverse Domains through World Models

World-Model. [ACM CSUR 2025] Understanding World or Predicting Future? A Comprehensive Survey of World Models

UnstableBaselines. Python

ml-aspera. Data generation engine, sandbox environment and evaluation tools accompanying "ASPERA: A Simulated Environment to Evaluate Planning for Complex Action Execution".

Awesome-RL-for-LRMs. A Survey of Reinforcement Learning for Large Reasoning Models

mlx. MLX: An array framework for Apple silicon

Automodel. 🚀 Pytorch Distributed native training library for LLMs/VLMs with OOTB Hugging Face support

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

sglang. SGLang is a high-performance serving framework for large language models and multimodal models.

TensorRT-LLM. TensorRT LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and supports state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs. TensorRT LLM also contains components to create Python and C++ runtimes that orchestrate the inference execution in a performant way.

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

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

RLVR-World. Official repository for "RLVR-World: Training World Models with Reinforcement Learning" (NeurIPS 2025), https://arxiv.org/abs/2505.13934

lmms-eval. One-for-All Multimodal Evaluation Toolkit Across Text, Image, Video, and Audio Tasks

Awesome-ML-SYS-Tutorial. My learning notes for ML SYS.

zed-mojo. Mojo support for Zed

materl. Python

aira-dojo. AIRA-dojo: a framework for developing and evaluating AI research agents

mamba. Mamba SSM architecture

czkawka. Multi functional app to find duplicates, empty folders, similar images etc.

mbridge. Bridge Megatron-Core to Hugging Face/Reinforcement Learning

kushim. eval creator

flash-linear-attention. 🚀 Efficient implementations for emerging model architectures

text-to-lora. Hypernetworks that adapt LLMs for specific benchmark tasks using only textual task description as the input

ROLL. An Efficient and User-Friendly Scaling Library for Reinforcement Learning with Large Language Models

horizon-reduction. The official implementation of "Horizon Reduction Makes RL Scalable"

ChatLearn. A flexible and efficient training framework for large-scale alignment tasks

MARTI. A Framework for LLM-based Multi-Agent Reinforced Training and Inference

modular. The Modular Platform (includes MAX & Mojo)

pi-quant. SIMD quantization kernels

verl-agent. verl-agent is an extension of veRL, designed for training LLM/VLM agents via RL. verl-agent is also the official code for paper "Group-in-Group Policy Optimization for LLM Agent Training"

trl. Train transformer language models with reinforcement learning.

verdict. Inference-time scaling for LLMs-as-a-judge.

inferx. InferX: Inference as a Service Platform

RL. Scalable toolkit for efficient model reinforcement

retrain. a Python library that uses Reinforcement Learning (RL) to train LLMs.

gg-bench. Measuring General Intelligence With Generated Games (Preprint)

SkyRL. SkyRL: A Modular Full-stack RL Library for LLMs

LlamaFactory. Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)

arbor. A framework for optimizing DSPy programs with RL

ReTool. Python

evorl. EvoRL is a fully GPU-accelerated framework for Evolutionary Reinforcement Learning, implemented with JAX. It supports Reinforcement Learning (RL), Evolutionary Computation (EC), Evolution-guided Reinforcement Learning (ERL), AutoRL, and seamless integration with GPU-optimized simulation environments.

awesome-RecSys. A curated list of awesome Recommender System (Books, Conferences, Researchers, Papers, Github Repositories, Useful Sites, Youtube Videos)

GRPO-Zero. Implementing DeepSeek R1's GRPO algorithm from scratch

cherry-studio. AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs

OpenHands. 🙌 OpenHands: AI-Driven Development

jiki. Python

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

ember. Python

griffe. Signatures for entire Python programs. Extract the structure, the frame, the skeleton of your project, to generate API documentation or find breaking changes in your API.

mkdocs-material. Documentation that simply works

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

lance. Open Lakehouse Format for Multimodal AI. Convert from Parquet in 2 lines of code for 100x faster random access, vector index, and data versioning. Compatible with Pandas, DuckDB, Polars, Pyarrow, and PyTorch with more integrations coming..

textbook-to-rl. Python

ReCall. ReSearch: Learning to Reason with Search for LLMs via Reinforcement Learning & ReCall: Learning to Reason with Tool Call for LLMs via Reinforcement Learning

verifiers. Our library for RL environments + evals

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

skrl. Modular Reinforcement Learning (RL) library (implemented in PyTorch, JAX, and NVIDIA Warp) with support for Gymnasium/Gym, NVIDIA Isaac Lab, MuJoCo Playground and other environments

pyreft. Stanford NLP Python library for Representation Finetuning (ReFT)

HEXA-MoE. Official code for the paper "HEXA-MoE: Efficient and Heterogeneous-Aware MoE Acceleration with Zero Computation Redundancy"

pytype. A static type analyzer for Python code

hapax-py. A type-safe graph execution framework built on top of OpenLit for LLM pipelines

AI_Diplomacy. Frontier Models playing the board game Diplomacy.

klarity. See Through Your Models

parlant. Build reliable customer-facing AI agents with Parlant: an interaction control harness optimized for controlled, consistent, and predictable LLM interactions.

swag. Automatically generate RESTful API documentation with Swagger 2.0 for Go.

beszel. Lightweight server monitoring with historical data, docker stats, and alerts.

quic-go. A production-ready QUIC implementation in pure Go

hapax. The reliability layer between your code and LLM providers.

fire. A lightweight, agent-style framework for fact-checking atomic claims using iterative retrieval and verification. Reduces LLM and search cost while maintaining strong factuality performance.

ghq. Remote repository management made easy

freeze. Generate images of code and terminal output 📸

deepflow. eBPF Observability - Distributed Tracing and Profiling

superfile. Pretty fancy and modern terminal file manager

raggo. A lightweight, production-ready RAG (Retrieval Augmented Generation) library in Go.

enzu-go. A framework for building multi-agent AI systems. Enables LLMs to collaborate through hierarchical organization, parallel task execution, and extensible tools.

goship. Opinionated Go + HTMX framework for shipping production apps fast ⛵️

UniversalBackrooms. A replication of Andy Ayrey's "Backrooms" (https://dreams-of-an-electric-mind.webflow.io/), but runnable with Opus 3, Sonnet 3.5, GPT 4o, and o1-preview

gobot. Golang framework for robotics, drones, and the Internet of Things (IoT)

Awesome-Out-Of-Distribution-Detection. Out-of-distribution detection, robustness, and generalization resources. The repository contains a curated list of papers, tutorials, books, videos, articles and open-source libraries etc

bridget. Bridget is a minimal Hugo theme for photographers/visual artists, powered by SolidJS.

operatio. Mojo

inferencemax.

Localize-and-Stitch. Localize-and-Stitch: Efficient Model Merging via Sparse Task Arithmetic

hyperobject. Plotting (entropy, varentropy) for small LMs

swarm. Educational framework exploring ergonomic, lightweight multi-agent orchestration. Managed by OpenAI Solution team.

gosql. A SQL query engine in Go

dicedb. Open-source, low-latency key/value engine built on Valkey with query subscriptions and hierarchical storage tiers.

zep. Zep | Examples, Integrations, & More

gofr. An opinionated GoLang framework for accelerated microservice development. Built in support for databases and observability.

sioyek. Sioyek is a PDF viewer with a focus on textbooks and research papers

entropix. Entropy Based Sampling and Parallel CoT Decoding

oscar. [mirror] Open source contributor agent architecture repo.

antislop-sampler. Python

SLOP_Detector. SLOP Detector and analyzer based on dictionary for shareGPT JSON and text

parler-tts. Inference and training library for high-quality TTS models.

faster-whisper. Faster Whisper transcription with CTranslate2

textual. The lean application framework for Python. Build sophisticated user interfaces with a simple Python API. Run your apps in the terminal and a web browser.

mesop. Rapidly build AI apps in Python

GPU-Puzzles. Solve puzzles. Learn CUDA.

matmulfreellm. Implementation for MatMul-free LM.

evennia. Python MUD/MUX/MUSH/MU* development system

zipnn. A Lossless Compression Library for AI pipelines

mojo-learning. 📖 Learn some mojo !

WildBench. Benchmarking LLMs with Challenging Tasks from Real Users

harbor. Stop configuring your AI stack. Start using it. One command brings a complete pre-wired LLM stack with hundreds of services to explore.

MisguidedAttention. A collection of prompts to challenge the reasoning abilities of large language models in presence of misguiding information

klmbr. klmbr - a prompt pre-processing technique to break through the barrier of entropy while generating text with LLMs

Awesome-Interpretability-in-Large-Language-Models. This repository collects all relevant resources about interpretability in LLMs

zml. Any model. Any hardware. Zero compromise. Built with @ziglang / @openxla / MLIR / @bazelbuild

gjson. Get JSON values quickly - JSON parser for Go

g1. g1: Using Llama-3.1 70b on Groq to create o1-like reasoning chains

Multi-Agents-Debate. MAD: The first work to explore Multi-Agent Debate with Large Language Models :D

colly. Elegant Scraper and Crawler Framework for Golang

go. A high-performance 100% compatible drop-in replacement of "encoding/json"

air. ☁️ Live reload for Go apps

flatbuffers. FlatBuffers: Memory Efficient Serialization Library

good-ui. My summary of the book "Refactoring UI" by Steve Schoger and Adam Wathan

SuperPrompt. SuperPrompt is an attempt to engineer prompts that might help us understand AI agents.

pdfcpu. PDF tooling for Go and the command line.

yousim. they've simulated websites, worlds, and imaginary CLIs... but what if they simulated *you*?

avante.nvim. Use your Neovim like using Cursor AI IDE!

open-resume. OpenResume is a powerful open-source resume builder and resume parser. https://open-resume.com/

semantic-grep. grep for words with similar meaning to the query

chromem-go. Embeddable vector database for Go with Chroma-like interface and zero third-party dependencies. In-memory with optional persistence.

geziyor. Geziyor, blazing fast web crawling & scraping framework for Go. Supports JS rendering.

form. Easily create HTML forms with Go structs.

gofh. GoFastHTML

Trace. End-to-end Generative Optimization for AI Agents

pocketbase. Open Source realtime backend in 1 file

cyclops. Developer Friendly Kubernetes 👁️

gonb. GoNB, a Go Notebook Kernel for Jupyter

jsonschema. Generate JSON Schemas from Go types

env. A simple, zero-dependencies library to parse environment variables into structs

linkedin-jobs-scraper. TypeScript

goth. Package goth provides a simple, clean, and idiomatic way to write authentication packages for Go web applications.

gollm. Unified Go interface for Language Model (LLM) providers. Simplifies LLM integration with flexible prompt management and common task functions.

cascade. Kubernetes native Chaos Engineering Toolkit | Reimplemented Netflix's Chaos Monkey

daytona. Daytona is a Secure and Elastic Infrastructure for Running AI-Generated Code

microservices-demo. Sample cloud-first application with 10 microservices showcasing Kubernetes, Istio, and gRPC.

garak. the LLM vulnerability scanner

Python. All Algorithms implemented in Python

system-design-primer. Learn how to design large-scale systems. Prep for the system design interview. Includes Anki flashcards.

semantic-router. Superfast AI decision making and intelligent processing of multi-modal data.

screenshot-to-code. Drop in a screenshot and convert it to clean code (HTML/Tailwind/React/Vue)

taipy. Turns Data and AI algorithms into production-ready web applications in no time.

AgentForge. Extensible AGI Framework

autogen. A programming framework for agentic AI

Adala. Adala: Autonomous DAta (Labeling) Agent framework

awesomeMLSys. An ML Systems Onboarding list