The first continuous diffusion language model that rivals discrete counterparts on standard language modeling benchmarks like LM1B and OpenWebText.

dUltra: Ultra-Fast Diffusion Large Language Models via Reinforcement Learning

Real-time global intelligence dashboard. AI-powered news aggregation, geopolitical monitoring, and infrastructure tracking in a unified situational awareness interface

Build RL environments for LLM training

[ICLRW'26] EoRA: Fine-tuning-free Compensation for Compressed LLM with Eigenspace Low-Rank Approximation

[Neurips 2025] R-KV: Redundancy-aware KV Cache Compression for Reasoning Models

Browser extension to be more focused on Youtube. 15000+ users. Supports Chrome, Firefox, Brave and Edge.

Anki plugin to expose a remote API for creating flash cards.

Scalable toolkit for efficient model reinforcement

Code for accepted paper at ICLR 2026

Why Diffusion Language Models Struggle with Truly Parallel (Non-Autoregressive) Decoding?

Thorsten-Voice: A free to use, offline working, high quality german TTS voice should be available for every project without any license struggling.

German Tacotron 2 and Multi-band MelGAN in TensorFlow with TF Lite inference support

DeepEP: an efficient expert-parallel communication library

A framework for the evaluation of autoregressive code generation language models.

SGLang is a fast serving framework for large language models and vision language models.

A project to improve skills of large language models

DFlash: Block Diffusion for Flash Speculative Decoding

ToolOrchestra is an end-to-end RL training framework for orchestrating tools and agentic workflows.

utilities for decoding deep representations (like sentence embeddings) back to text

Code for NeurIPS 2024 paper - The GAN is dead; long live the GAN! A Modern Baseline GAN - by Huang et al.

Official JAX implementation of End-to-End Test-Time Training for Long Context

[ICLR 2026] Official PyTorch Implementation of RLP: Reinforcement as a Pretraining Objective

dInfer: An Efficient Inference Framework for Diffusion Language Models

GPU-optimized framework for training diffusion language models at any scale. The backend of Quokka, Super Data Learners, and OpenMoE 2 training.

CANDI: Continuous and Discrete Diffusion

Leetcode for Pytorch

[NeurIPS'19] Deep Equilibrium Models

🧀 Pytorch code for the Fromage optimiser.

Code for 'LLM2Vec: Large Language Models Are Secretly Powerful Text Encoders'