Stars
Tile primitives for speedy kernels
Accelerating the development of large multimodal models (LMMs) with one-click evaluation module - lmms-eval.
Evaluating Multimodal Generative AI with Korean Educational Standards, NAACL 2025.
A generative world for general-purpose robotics & embodied AI learning.
Acceleration framework for Human Alignment Learning
Official PyTorch implementation of “MaskRIS: Semantic Distortion-aware Data Augmentation for Referring Image Segmentation”
Data processing for and with foundation models! 🍎 🍋 🌽 ➡️ ➡️🍸 🍹 🍷
A suite of image and video neural tokenizers
Easily turn large sets of image urls to an image dataset. Can download, resize and package 100M urls in 20h on one machine.
MambaOut: Do We Really Need Mamba for Vision? (CVPR 2025)
Line-granularity, thread-aware deterministic and statistic pure-python profiler
[ICLR 2024] MogaNet: Efficient Multi-order Gated Aggregation Network
[CVPR 2024 Best paper award candidate] EGTR: Extracting Graph from Transformer for Scene Graph Generation
[ECCV 2024] Official PyTorch implementation of RoPE-ViT "Rotary Position Embedding for Vision Transformer"
Model Stock: All we need is just a few fine-tuned models
[ECCV2024] Official implementation of paper, "DenseNets Reloaded: Paradigm Shift Beyond ResNets and ViTs".
Interactive textbook on state-space models
The WeightWatcher tool for predicting the accuracy of Deep Neural Networks
[CVPR 2024] Code release for TransNeXt model