William Luo's repositories
diff_instruct
official code for Diff-Instruct algorithm for one-step diffusion distillation
3DFuse
Official implementation of "Let 2D Diffusion Model Know 3D-Consistency for Robust Text-to-3D Generation"
3dgp
3D generation on ImageNet [ICLR 2023]
awesome-demos
links and status of cool gradio demos
BLIP
PyTorch code for BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation
Coin-SVGD
Code for "Coin Sampling: Gradient-Based Bayesian Inference without Learning Rates" (ICML 2023)
consistency_models
Official repo for consistency models.
dalle-mini
DALL·E Mini - Generate images from a text prompt
DG
Official repo for Discriminator Guidance.
DG_imagenet
Official repo for Discriminator Guidance for ImageNet256.
edit3d
Code for Cross-Modal 3D Shape Generation and Manipulation (ECCV 2022)
fast-stable-diffusion
fast-stable-diffusion + DreamBooth
Fengshenbang-LM
Fengshenbang-LM(封神榜大模型)是IDEA研究院认知计算与自然语言研究中心主导的大模型开源体系,成为中文AIGC和认知智能的基础设施。
flash-attention
Fast and memory-efficient exact attention
kaolin
A PyTorch Library for Accelerating 3D Deep Learning Research
mae
PyTorch implementation of MAE https//arxiv.org/abs/2111.06377
minddiffusion
A collection of diffusion models based on MindSpore
NeuralMJP
Code accompanying the paper "Neural Markov Jump Processes" published at ICML 2023.
open_clip
An open source implementation of CLIP.
ParEM
Code for "Particle algorithms for maximum likelihood training of latent variable models" (Kuntz, Lim, Johansen, AISTATS, 2023).
pkulwj1994.github.io
Weijian Luo homepage
poisson-jump
Official Implementation of Paper "Learning to Jump: Thinning and Thickening Latent Counts for Generative Modeling" (ICML 2023)
RCFD
[ICME-2023] Official implementation for "Accelerating Diffusion Sampling with Classifier-based Feature Distillation".
RWKV-LM
RWKV is an RNN with transformer-level LLM performance. It can be directly trained like a GPT (parallelizable). So it's combining the best of RNN and transformer - great performance, fast inference, saves VRAM, fast training, "infinite" ctx_len, and free sentence embedding.
shap-e
Generate 3D objects conditioned on text or images
stylegan-xl
[SIGGRAPH'22] StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasets
threestudio
A unified framework for 3D content generation.