The author's implementation of FUDOKI, a multimodal large language model purely based on discrete flow matching.

Long-RL: Scaling RL to Long Sequences

Locality-aware Parallel Decoding for Efficient Autoregressive Image Generation

Official implementation of "Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decoding"

Official PyTorch implementation for "Large Language Diffusion Models"

New repo collection for NVIDIA Cosmos: https://github.com/nvidia-cosmos

SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformer

DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding

SEED-Voken: A Series of Powerful Visual Tokenizers

Minimal implementation of scalable rectified flow transformers, based on SD3's approach

Next-Token Prediction is All You Need

Janus-Series: Unified Multimodal Understanding and Generation Models

[TMLR 2025🔥] A survey for the autoregressive models in vision.

An open source implementation of CLIP.

Autoregressive Model Beats Diffusion: 🦙 Llama for Scalable Image Generation

Code and models for ICML 2024 paper, NExT-GPT: Any-to-Any Multimodal Large Language Model

Adapting LLaMA Decoder to Vision Transformer

Taming Transformers for High-Resolution Image Synthesis

PyTorch package for the discrete VAE used for DALL·E.

The official Meta Llama 3 GitHub site

Grok open release

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

[ICML'24 Spotlight] LLM Maybe LongLM: Self-Extend LLM Context Window Without Tuning

MetaMath: Bootstrap Your Own Mathematical Questions for Large Language Models

Tools for merging pretrained large language models.

Example models using DeepSpeed