Current Behavior:

Currently, the filter engine in our travel planning system is deleting some days from the travel plan, which may result in an incomplete itinerary for the users.

Expected Behavior:

We need to implement a new engine that calls the Large Language Model (LLM) to complete the travel plan automatically. The goal is to ensure that the final itinerary matches the number of days required by the user, even if some days were removed by the filter engine.

Hey there 👋 Just wanted to let you know that your app on Streamlit Cloud deployed from this repo has gone over its resource limits. Access to the app is temporarily limited. Visit the app to see more details and possible solutions.

Fix security alerts:

• Upgrade aiohttp to version 3.8.5 or later.
• Upgrade cryptography to version 41.0.2 or later.

Objective:

The goal of this issue is to develop a prototype engine that calculates a proxy price for the travel plan generated by the LLM. The calculated price should primarily consider flight and accommodation costs. The engine should use a unified API for obtaining hotel and flight prices, starting with TripAdvisor as the initial choice. The ultimate target is to fetch flight prices from Skyscanner and hotel prices from Booking.com and Airbnb. Additionally, the prototype should display links to the selected flights and hotels on the frontend of the travel plan.

Features to Implement:

• Proxy Price Calculation: Develop a mechanism within the engine that computes an estimated cost for the travel plan based on the flight and accommodation expenses.

• Unified API Integration: Initially, integrate the engine with the TripAdvisor API to retrieve hotel and flight prices. The engine should be designed with flexibility to later incorporate Skyscanner API for flights and Booking.com and Airbnb APIs for hotels.

• Flight Price Retrieval: Implement functionality to fetch flight prices from the appropriate API (e.g., Skyscanner) based on the user's specified travel dates and destinations.

• Hotel Price Retrieval: Implement functionality to retrieve hotel prices from the selected API(s) (e.g., Booking.com and Airbnb) based on the user's desired accommodation options.

• Frontend Link Generation: Ensure that the prototype generates relevant links to the chosen flights and hotels on the travel plan frontend. These links will provide users with direct access to booking options.

Current Status:

Currently, the app displays the labels of travel activities that have a probability greater than 0.5, which is functioning as intended.

Expected Behavior:

However, when a travel activity does not have any label with a probability greater than 0.5, the app should print the argmax label instead to provide more accurate information to the users.

Hi, I'm trying to test the app on streamlit, but I get always:
"Not valid OpenAI API Access Key"

I tried with both types of OpenAI keys (user and project).
Am I doing something wrong?