Stars
100M tokens. Infinite compute. Lowest val loss wins.
Fast, Sharp & Reliable Agentic Intelligence
Step3-VL-10B: A compact yet frontier multimodal model achieving SOTA performance at the 10B scale, matching open-source models 10-20x its size.
STEP-GUI: The top GUI agent solution in the galaxy. Developed by the StepFun-GELab team and powered by StepFun’s cutting-edge research capabilities.
Everything about the SmolLM and SmolVLM family of models
Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels
专门用于给图片加水印打码的工具，完全基于浏览器本地API，无任何网络请求（特别适合身份证等敏感证件）
A PyTorch library for implementing flow matching algorithms, featuring continuous and discrete flow matching implementations. It includes practical examples for both text and image modalities.
[CVPR 2025 Oral] Reconstruction vs. Generation: Taming Optimization Dilemma in Latent Diffusion Models
A robust, efficient, low-latency speech-to-text library with advanced voice activity detection, wake word activation and instant transcription.
[ICML2025] Make LoRA Great Again: Boosting LoRA with Adaptive Singular Values and Mixture-of-Experts Optimization Alignment
A simple, modern and secure encryption tool (and Go library) with small explicit keys, no config options, and UNIX-style composability.
[COLM 2025] An Open Math Pre-trainng Dataset with 370B Tokens.
Chinese-Vicuna: A Chinese Instruction-following LLaMA-based Model —— 一个中文低资源的llama+lora方案，结构参考alpaca
[NeurIPS2024] Twin-Merging: Dynamic Integration of Modular Expertise in Model Merging
Powerline is a statusline plugin for vim, and provides statuslines and prompts for several other applications, including zsh, bash, tmux, IPython, Awesome and Qtile.
Byted PyTorch Distributed for Hyperscale Training of LLMs and RLs
Seas0 / libcimbar
Forked from sz3/libcimbarOptimized implementation for color-icon-matrix barcodes
SGLang is a high-performance serving framework for large language models and multimodal models.
😎 A Survey of Efficient Reasoning for Large Reasoning Models: Language, Multimodality, Agent, and Beyond
[EMNLP2023]: MIRACLE: Towards Personalized Dialogue Generation with Latent-Space Multiple Personal Attribute Control
[CVPR2025] Extrapolating and Decoupling Image-to-Video Generation Models: Motion Modeling is Easier Than You Think
🚀 LLaMA-MoE v2: Exploring Sparsity of LLaMA from Perspective of Mixture-of-Experts with Post-Training