CoolFace
Modelpublic

UI-Simulator/UI_Simulator_Grow_Android_Step1

sourceHugging Facemitupdated 11mo agoView on Hugging Face
0likes11downloads
README.md40 linesDownload Raw Back to root
1---2license: mit3datasets:4- UI-Simulator/UI-Simulator_android_data5language:6- en7base_model:8- Qwen/Qwen2.5-7B-Instruct9pipeline_tag: text-generation10tags:11- agent12- mobile-agent13- world-model14- GUI-agent15---16 17# ๐ŸŒ UI-Simulator: LLMs as Scalable, General-Purpose Simulators For Evolving Digital Agent Training18 19[![License](https://img.shields.io/github/license/WadeYin9712/UI-Simulator)](LICENSE)20[![Python](https://img.shields.io/badge/python-3.8%2B-blue.svg)]()21[![Build](https://img.shields.io/badge/build-passing-brightgreen.svg)]()22[![Stars](https://img.shields.io/github/stars/WadeYin9712/UI-Simulator?style=social)]()23 24๐Ÿ”— **Paper link**: [Link]() 25 26๐ŸŒ **Website**: https://ui-simulator.notion.site/llms-as-scalable-digital-world-simulator27 28๐Ÿ“š **Github:** https://github.com/WadeYin9712/UI-Simulator29 30๐Ÿ“ง **Contact:** [da.yin9712@gmail.com](mailto:da.yin9712@gmail.com), [w10y20ming@gmail.com](mailto:w10y20ming@gmail.com)31 32---33 34# Model Card for Model ID35 36This is the Step1 agent trained with **retrieval-augmented** UI-Simulator with **UI-Simulator-Grow scaling paradigm**. It is trained to solve AndroidWorld tasks.37 38## Uses39 40We provide a script for loading the model and run AndroidWorld evaluation [here](https://github.com/WadeYin9712/UI-Simulator/blob/main/android_world/run.py).