add docs description

app.py

@@ -155,6 +155,10 @@ Using this model, we can index images and perform text-to-image retrieval.
 
 This demo compares to pure text retrieval using the `BAAI/bge-small-en` model. Is this a fair comparison? Not really,
 but it's the easiest to run in a limited huggingface space, and shows the strengths of screenshot-based retrieval.
+                
+The two pre-made indexes are:
+- [IONIQ 2024](https://cdn.prod.website-files.com/65e1d7fb19a3e64b5c36fb38/66eb856e019e59758ef73759_ICONIQ%20Analytics%20%2B%20Insights%20-%20State%20of%20AI%20Sep24.pdf): A report on the 2024 State of AI.
+- [Uber 10k 2021](https://github.com/run-llama/llama_index/blob/main/docs/docs/examples/data/10k/uber_2021.pdf): The 2021 Uber 10k document.
 """
     )