Specialized Domains

High-performance object detection library.

Stanford NLP Python library for 100+ human languages. State-of-the-art neural pipelines for tokenization, NER, parsing, and sentiment analysis with pre-trained models. Apache 2.0 licensed.

World's most widely used computer vision library.

State-of-the-art real-time object detection.

Global weather and tropical cyclone forecasting framework from Google DeepMind, including WeatherNext 2, GraphCast, and GenCast models. Apache 2.0 licensed.

Open-source deep learning framework for physics-informed machine learning (Physics-ML). Build, train, and fine-tune models for AI4science and engineering applications using state-of-the-art SciML methods. Apache 2.0 licensed.

PyTorch domain library for geospatial data. Datasets, samplers, transforms, and pre-trained models for multispectral satellite imagery and remote sensing. First library with pre-trained models for Sentinel-2 bands. MIT licensed.

Core library for astronomy and astrophysics in Python. Comprehensive tools for celestial coordinates, FITS I/O, cosmological calculations, and data analysis for professional astronomy. BSD-3-Clause licensed.

Open-source biomolecular interaction prediction models. Boltz-1 was the first fully open source model to approach AlphaFold3 accuracy; Boltz-2 adds binding affinity prediction for drug discovery. MIT licensed.

High-accuracy open-source biomolecular structure prediction model from ByteDance. First fully open-source model to outperform AlphaFold3 across diverse benchmarks with Apache 2.0 licensing for both academic and commercial use.

Democratizing deep learning for drug discovery, quantum chemistry, materials science, and biology. High-quality open-source toolchain with 50+ models and extensive tutorials. MIT licensed.

Modern, comprehensive probabilistic programming framework in Python. Bayesian modeling with advanced MCMC sampling, variational inference, and seamless integration with ArviZ for visualization. Apache 2.0 licensed.

Exploratory analysis of Bayesian models with Python. Comprehensive visualization and diagnostics for probabilistic models, supporting PyMC, Pyro, Stan, and other PPLs. Apache 2.0 licensed.

Medical Open Network for AI. End-to-end framework for healthcare imaging with state-of-the-art, production-ready training workflows. Apache 2.0 licensed.

Self-configuring deep learning method for medical image segmentation. Automatically adapts to any dataset without manual parameter tuning. Widely adopted as the standard baseline for biomedical segmentation challenges. Apache 2.0 licensed.

Local-first healthcare AI framework and application for clinical text de-identification, entity extraction, and on-device specialized model serving. Apache 2.0 licensed.

Open-source desktop novel-writing engine and AI story scaffold with Tauri, FastAPI, SQLite, and plot-generation workflows. MIT licensed.

Local-first desktop workbench for long-form fiction that organizes story premises, characters, worldbuilding, chapter blueprints, drafting, review, and revision, with Ollama support and Windows/macOS releases. GPL-3.0 licensed.

Reproducible training recipe and open weights for a Mistral-Nemo-12B creative-writing fine-tune built with QLoRA and anti-slop DPO, with GGUF quants and the lowest slop score among the same-size 12B models tested on EQ-Bench Creative Writing v3. Apache-2.0 licensed.

Self-hostable AI interview platform for voice, chat, and video interviews with adaptive follow-ups, automated scoring, and detailed feedback reports. MIT licensed.

Toolkit for training intelligent agents in games and simulations using deep reinforcement learning. Enables NPC behavior control, automated testing, and game design evaluation. Apache 2.0 licensed.

A training framework for Stable Baselines3 reinforcement learning agents with hyperparameter optimization, pre-trained agents, and extensive benchmark environments. MIT licensed.

Modular reinforcement learning library implemented in PyTorch, JAX, and NVIDIA Warp with support for Gymnasium, NVIDIA Isaac Lab, MuJoCo Playground, and other environments. MIT licensed.

Scalable and memory-optimized training of diffusion models from Hugging Face. Supports LoRA and full fine-tuning for video and image generation models. Apache 2.0 licensed.