A minimal implementation of DeepMind's Genie world model

PyTorch code and models for VJEPA2 self-supervised learning from video.

DIAMOND (DIffusion As a Model Of eNvironment Dreams) is a reinforcement learning agent trained in a diffusion world model. NeurIPS 2024 Spotlight.

Can the V-JEPA2 model be used as a world model?

PyTorch code and models for VJEPA2 self-supervised learning from video.

Tiny AutoEncoder for Stable Diffusion (and other image models)

Implementation of a JEPA Image World Model, trained on OpenAI's VPT Minecraft contractor dataset.

A carefully curated collection of high-quality tools, libraries, research papers, projects, and tutorials centered around Joint Embedding Predictive Architecture (JEPA).

This is the official repository for the paper: JEPA-T: Joint-Embedding Predictive Architecture with Text Fusion for Image Generation

jepa but nanogpt style

Official repository for TransGeo: Transformer Is All You Need for Cross-view Image Geo-localization

[CVPR 2026] CFG-Ctrl: Control-Based Classifier-Free Diffusion Guidance

Fast, small, and fully autonomous AI personal assistant infrastructure, ANY OS, ANY PLATFORM — deploy anywhere, swap anything 🦀

⚡️ Lightning-fast backtesting engine to find your trading edge.

Modular reinforcement learning framework for algorithmic trading

Tiny, Fast, and Deployable anywhere — automate the mundane, unleash your creativity

Because `model.fit()` isn't an explanation

Free audio transcript generator, running locally in the browser

An open-source audio wake word (or phrase) detection framework with a focus on performance and simplicity.

An unofficial Pinterest media downloader

Influencer dataset collected from Instagram

Andrej Karpathy's micrograd library implemented in Go

Qwen-Image is a powerful image generation foundation model capable of complex text rendering and precise image editing.

cDCGAN for generating handwritten text

[SIGGRAPH'22] StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasets

Implementation of 'lightweight' GAN, proposed in ICLR 2021, in Pytorch. High resolution image generations that can be trained within a day or two