🏠[ECCV 2024] GaussianImage: 1000 FPS Image Representation and Compression by 2D Gaussian Splatting

MiniCPM-2B: An end-side LLM outperforming Llama2-13B.

The hub for EleutherAI's work on interpretability and learning dynamics

ACE0 is a learning-based structure-from-motion approach that estimates camera parameters of sets of images by learning a multi-view consistent, implicit scene representation.

The repo for "Metric3D: Towards Zero-shot Metric 3D Prediction from A Single Image" and "Metric3Dv2: A Versatile Monocular Geometric Foundation Model..."

Code for PhysDreamer

[Arxiv] A Survey on Video Diffusion Models

an unofficial 2DGS implementation based on GauStudio

The official Meta Llama 3 GitHub site

CLAY: A Controllable Large-scale Generative Model for Creating High-quality 3D Assets

LLM training in simple, raw C/CUDA

The simplest, fastest repository for training/finetuning medium-sized GPTs.

🚀 跃问YueWen 多模态大模型逆向API白嫖测试【特长：超强多模态】，支持高速流式输出、联网搜索、长文档解读、图像解析、多轮对话，零配置部署，多路token支持，自动清理会话痕迹。

Code for ICLR 2024 paper "Duolando: Follower GPT with Off-Policy Reinforcement Learning for Dance Accompaniment"

Open source implementation and models of One-step Diffusion with Distribution Matching Distillation

DUSt3R: Geometric 3D Vision Made Easy

[CVPR 2024 Highlight] DistriFusion: Distributed Parallel Inference for High-Resolution Diffusion Models

High-quality multi-lingual text-to-speech library by MyShell.ai. Support English, Spanish, French, Chinese, Japanese and Korean.

[CVPR 2024 - Oral, Best Paper Award Candidate] Marigold: Repurposing Diffusion-Based Image Generators for Monocular Depth Estimation

This is the official code release for our work, Denoising Vision Transformers.

News: the 7k dataset is ready for download.

A toolkit for 3D computer vision tasks.

Emu Series: Generative Multimodal Models from BAAI

An open-source impl. of Large Reconstruction Models

Machine Learning Engineering Open Book

Track-Anything is a flexible and interactive tool for video object tracking and segmentation, based on Segment Anything, XMem, and E2FGVI.

4D-fy: Text-to-4D Generation Using Hybrid Score Distillation Sampling