Update README.md

README.md

@@ -1,3 +1,28 @@
----
-license: mit
----
+---
+license: mit
+---
+# DEVAI dataset
+<p align="center" width="100%">
+<img src="dataset_stats.png" align="center" width="70%"/>
+</p>
+
+**DEVAI** is a benchmark of 55 realistic AI development tasks. It consists of plentiful manual annotations, including a total of 365 hierarchical user requirements. 
+This dataset enables rich reinforcement signals for better automated AI software development. 
+
+Here is an example of our tasks.
+<p align="center" width="100%">
+<img src="task51.png" align="center" width="60%"/>
+</p>
+
+We apply three state-of-the-art automatic software development systems to DEVAI, namely MetaGPT, GPT-Piolt, and OpenHands. The table below shows preliminary statistics results.
+<p align="center" width="100%">
+<img src="developer_stats.png" align="center" width="60%"/>
+</p>
+
+We perform a manual evaluation to judge if each requirement is satisfied by the solution provided by the aforementioned systems.
+<p align="center" width="100%">
+<img src="human_evaluation.png" align="center" width="60%"/>
+</p>
+
+An automated evaluation program that could possibly replace manual evaluation can be found at our [Github realse](https://github.com/metauto-ai/Devai).
+Find more details in our [paper]().