I am trying to write a streamlit app that takes a prompt, takes a photo, and then run a GPT-4 Vision analysis on the photo. The app does not run beyond line ~13 (the screen just stops after the photo is taken, with no traceback). Would anyone know what the issue might be?