2U1
Follow
Stars
An open-source implementaion for Gemma3 series by Google.
This repository aims to develop CoT Steering based on CoT without Prompting. It focuses on enhancing the model’s latent reasoning capability without additional training by leveraging Test-Time Scal…
✨First Open-Source R1-like Video-LLM [2025/02/18]
[ICLR 2025] Alignment Data Synthesis from Scratch by Prompting Aligned LLMs with Nothing. Your efficient and high-quality synthetic data generation pipeline!
A server application designed on top of MCP to interact with Cursor and MySQL.
Official code repository for Sketch-of-Thought (SoT)
No fortress, purely open ground. OpenManus is Coming.
Applied AI experiments and examples for PyTorch
[CVPR 2025] Official implementation of the paper "DiGIT: Multi-Dilated Gated Encoder and Central-Adjacent Region Integrated Decoder for Temporal Action Detection Transformer"
DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling
Muon optimizer: +>30% sample efficiency with <3% wallclock overhead
Frontier Multimodal Foundation Models for Image and Video Understanding
The official implementation of "Adapter is All You Need for Tuning Visual Tasks".
Solve Visual Understanding with Reinforced VLMs
lightweight, standalone C++ inference engine for Google's Gemma models.
An algorithm for weight-activation quantization (W4A4, W4A8) of LLMs, supporting both static and dynamic quantization
This repo contains the code for "VLM2Vec: Training Vision-Language Models for Massive Multimodal Embedding Tasks" [ICLR25]
A RLHF Infrastructure for Vision-Language Models
The official implementation of the paper "MMFuser: Multimodal Multi-Layer Feature Fuser for Fine-Grained Vision-Language Understanding". MMFuser addresses the limitations of current MLLMs in captur…
A modular, primitive-first, python-first PyTorch library for Reinforcement Learning.
A Flexible Framework for Experiencing Cutting-edge LLM Inference Optimizations