Improve UltraEditBench dataset card: Add metadata, GitHub link, and sample usage

This pull request significantly enhances the dataset card for UltraEditBench by:
- Adding `task_categories: ['question-answering', 'text-generation']`, `language: en`, and `tags: ['model-editing', 'lifelong-learning']` to the metadata for improved discoverability and categorization on the Hugging Face Hub.
- Including a direct link to the associated GitHub repository (`https://github.com/XiaojieGu/UltraEdit`) for easy access to the code.
- Adding a dedicated "Sample Usage" section, incorporating setup and run instructions with code snippets directly extracted from the project's GitHub README, to guide users on how to interact with the dataset and the UltraEdit framework.

The existing arXiv paper link has been preserved as per instructions.

README.md

@@ -1,3 +1,13 @@
+---
+task_categories:
+- question-answering
+- text-generation
+language: en
+tags:
+- model-editing
+- lifelong-learning
+---
+
 # UltraEditBench
 
 UltraEditBench is the largest publicly available dataset to date for the task of model editing. 
@@ -6,6 +16,8 @@ This dataset was introduced in the paper:
 
 > [ULTRAEDIT: Training-, Subject-, and Memory-Free Lifelong Editing in Large Language Models](https://arxiv.org/abs/2505.14679)
 
+Code: https://github.com/XiaojieGu/UltraEdit
+
 ---
 
 ## 📦 Dataset Overview
@@ -44,6 +56,39 @@ Each sample in UltraEditBench includes three full instances (question–answer p
 
 ---
 
+## 🚀 Sample Usage
+
+### Setup
+Create the environment and install dependencies:
+
+```bash
+conda create -n ultraedit python=3.10
+conda activate ultraedit
+pip install torch==2.3.0+cu121 --index-url https://download.pytorch.org/whl/cu121
+pip install -r requirements.txt
+```
+💡 If you want to try editing a Mistral-7B model, even a **24GB consumer GPU** is enough — model editing for everyone!
+
+### Run Experiments
+Run the main experiment with:
+
+```bash
+sh run.sh
+```
+
+The `run.sh` script includes a sample command like:
+
+```
+python main.py dataset=zsre model=mistral-7b editor=ultraedit num_seq=200 \ # Number of turns
+    editor.cache_dir=cache \
+    dataset.batch_size=10 \
+    dataset.n_edits=100 \ # Number of edits per turn
+    model.edit_modules="[model.layers.29.mlp.down_proj, model.layers.30.mlp.down_proj]"
+```
+💡 Just try editing **20K samples** on Mistral-7B in **under 5 minutes** — ultra-efficient!
+
+---
+
 ## 💡 Citation
 
 If you use this dataset, please cite:
@@ -60,4 +105,5 @@ If you use this dataset, please cite:
 ## 📨 Contact
 
 - **Email**: [peettherapynoys@gmail.com](mailto:peettherapynoys@gmail.com)
-- **GitHub Issues**: [github.com/XiaojieGu/UltraEdit](https://github.com/XiaojieGu/UltraEdit/issues)
+- **GitHub Issues**: [github.com/XiaojieGu/UltraEdit](https://github.com/XiaojieGu/UltraEdit/issues)
+```