Awesome Opensource Ai

Reference implementations of MLPerf® training benchmarks

MLE-bench is a benchmark for measuring how well AI agents perform at machine learning engineering

Synthetic data curation for post-training and structured data extraction

Lightweight and extensible compatibility layer between dataframe libraries!

OpenMLDB is an open-source machine learning database that provides a feature platform computing consistent features for training and inference.

Automated Machine Learning on Kubernetes

Machine learning with dataframes

A fast and lightweight framework for creating decentralized agents with ease.

A lightweight, low-dependency, unified API to use all common reranking and cross-encoder models.

Code repo for "WebArena: A Realistic Web Environment for Building Autonomous Agents"

Reference implementations of MLPerf® inference benchmarks

Fast Multimodal LLM on Mobile Devices

Training Sparse Autoencoders on Language Models

The LLM Anti-Framework

The open document intelligence platform for builders and hackers - DMS for the agentic world

bloom - evaluate any behavior immediately  🌸🌱

Scalable and memory-optimized training of diffusion models

🌎💪 BrowserGym, a Gym environment for web task automation

Ground truth layer for humans and AI agents working together. Version control for knowledge.

PinchBench is a benchmarking system for evaluating LLM models as OpenClaw coding agents. Made with 🦀 by the humans at https://kilo.ai

Ouroboros — self-creating AI agent. Born Feb 16, 2026.

Trained models & code to predict toxic comments on all 3 Jigsaw Toxic Comment Challenges. Built using ⚡ Pytorch Lightning and 🤗 Transformers. For access to our API, please email us at contact@unitary.ai.

Scalable machine 🤖 learning for time series forecasting.

XAI - An eXplainability toolbox for machine learning

🥤 RAGLite is a Python toolkit for Retrieval-Augmented Generation (RAG) with DuckDB or PostgreSQL

Readymade evaluators for your LLM apps

Lightning fast data version control system for large repositories of data. Feels like git, pushes and pulls like oxen.

NVTabular is a feature engineering and preprocessing library for tabular data designed to quickly and easily manipulate terabyte scale datasets used to train deep learning based recommender systems.

FAIR Sequence Modeling Toolkit 2

Weave is a toolkit for developing AI-powered applications, built by Weights & Biases.

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

Caura (formerly MemClaw) — governed shared memory for AI agent fleets. Multi-agent, multi-tenant, MCP-native. Trust tiers, keystone policies, audit trails, knowledge graph, self-improving retrieval. Apache 2.0.

Open-source memory runtime for AI agents — reproducible, provenance-tagged context bundles instead of query-time retrieval. Apache-2.0, self-hosted on Postgres + pgvector, Python + TypeScript SDKs.

OrcaReplay — Time travel for AI agents. Record, replay, fork, and debug any agent run with any model. Built by the OrcaRouter.ai team.

Unified CLI for running AI coding agents in isolated containers. Includes built-in local metrics collection, HTTP traffic tracking, and an analytics dashboard to track agent actions.

Self-Programming AI Agent Framework. Agents create and refine their own workflows.