Update README.md

README.md

@@ -1,23 +1,25 @@
----
-dataset_info:
-  features:
-  - name: id
-    dtype: string
-  - name: prompt
-    dtype: string
-  - name: end_time
-    dtype: string
-  - name: level
-    dtype: int64
-  splits:
-  - name: train
-    num_bytes: 57769
-    num_examples: 73
-  download_size: 17950
-  dataset_size: 57769
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
----
+# FutureX-Online Dataset
+
+Welcome to the **FutureX-Online** dataset, the official dataset for our weekly real-time prediction challenge!
+
+This dataset is a **live, weekly-updating collection of questions** designed to test your model's ability to predict the future. Every week, new questions are added, and the previous week's questions are archived.
+
+For access to all historical questions, please see our other dataset: [**futurex-ai/Futurex-Past**](https://huggingface.co/datasets/futurex-ai/Futurex-Past).
+
+To learn more about the FutureX project, visit our [**official website**](https://futurex-ai.github.io/).
+
+## 🎯 Join Our Weekly Prediction Challenge!
+
+Curious about how your model performs in predicting the actual future? Join our weekly real-time prediction challenge and compete against the world's top models!
+
+### How to Participate
+
+Participation is simple:
+
+1.  Just send an email to **FutureX-ai@outlook.com** and provide your model's API (must be compatible with the OpenAI API).
+2.  **Suggested Subject Line**: `Futurex Challenge Entry - [Your Model Name]`
+
+Upon receiving your submission, we will include your model in our weekly real-time evaluation. Every week, we will use the latest data from this dataset to make calls to your model, rank its performance, and display the results on a dedicated dynamic leaderboard for the challenge.
+
+We promise to keep your API Key confidential and use it only for competition evaluation. We look forward to your entry!
+