|
Download README.md from vagrt/MemU-Bench-data: direct link, hf CLI and curl.
- Browser
- Download file 836 Bytes
-
https://huggingface.co/datasets/vagrt/MemU-Bench-data/resolve/main/README.md
- Command line
-
hf download hf://datasets/vagrt/MemU-Bench-data/README.md
-
curl -L -o README.md https://huggingface.co/datasets/vagrt/MemU-Bench-data/resolve/main/README.md
836 Bytes
| license: apache-2.0 | |
| This repository contains the dataset for MemU-Bench, a benchmark for evaluating memory capabilities in GUI agents. | |
| The dataset includes 29 user interaction trajectories and their corresponding simulator databases: | |
| - **User Interaction Trajectories**: Interaction trajectories recorded as users perform tasks in the Android simulator. | |
| - **Simulator Databases**: The corresponding simulator databases collected after task execution, including information such as contacts, phone numbers, SMS messages, and other app data. | |
| Note: All data in the simulator, including contacts, phone numbers, SMS messages, and other personal information, is entirely fictional and synthetically generated. No real user data is included. | |
| The benchmark evaluation code is available at https://github.com/ZJU-REAL/MemU-Bench | |