Lucid Server: Backend for Lucid Journal

Lucid Server is the core backend for the Lucid Journal project, handling database functionality, API calls, AI logic, authorization, and more. Leveraging cutting-edge AI models like GPT-3.5/4 for text generation and DALLE 2 for image synthesis, Lucid Server brings dreams to life.

Features

• AI-Powered Text and Image Generation: Utilize GPT and DALLE models to generate dream-like text and images.
• RESTful API: Manage and analyze dream data with well-defined API endpoints.
• Flexible Deployment: Deploy locally or on Heroku with ease.

Related Projects

• Frontend: Lucid Journal Frontend Repository

Packages

Lucid Server utilizes several agent-centric Python packages developed by the Autonomous Research Group:

• agentmemory: Easy vectorDB using ChromaDB locally and Supabase Postgres in production.
• easycompletion: Streamlined OpenAI functions and validation.
• agentlogger: Simple and visually appealing logs for agents.

Local Setup

To set up the Lucid Journal backend server locally, follow these steps:

1. Install Python: Visit the official Python website and download the latest version of Python for your operating system.

2. Install the required dependencies using pip:

   pip install -r requirements.txt
   

3. Create a "config.ini" file: In the same directory as the API code, create a "config.ini" file and add the following content:

   [openai]
   api_key=YOUR_OPENAI_API_KEY
   

   Replace "YOUR_OPENAI_API_KEY" with your actual OpenAI API key.

Running the API with ngrok

To expose the API locally and access it from the Dream Guide app running on Expo, you can use ngrok. Follow these steps:

1. Install ngrok: Visit the ngrok website and download the appropriate version for your operating system.

2. In the terminal, navigate to the directory containing the Lucid Journal backend server.

3. Run the command to start ngrok and forward requests to the local API server on port 5000:

   ngrok http 5000
   

4. ngrok will generate a temporary public URL (e.g., http://xxxxxxx.ngrok.io) that forwards requests to your local API server.

API Endpoints

The Lucid Journal backend server provides several endpoints to interact with dream data and AI models. Some key endpoints include:

• POST /api/dreams: Create a new dream entry.
• PUT /api/dreams/{dream_id}: Update the analysis and image of a specific dream entry.
• GET /api/dreams: Get all saved dream entries.
• GET /api/dreams/{dream_id}: Get details of a specific dream entry.
• GET /api/dreams/{dream_id}/analysis: Get the analysis of a specific dream entry.
• GET /api/dreams/{dream_id}/image: Get the AI-generated dream-inspired image for a specific dream entry.
• POST /api/chat: Have interactive conversations with the AI dream guide.
• POST /api/dreams/search: Search for dream entries based on keywords.
• POST /api/dreams/search-chat: Have AI-guided conversations with the AI dream guide and relevant dream entries found in the database.
• more to be added soon: TODO: add all endpoints

Running the API

To start the API server, execute the following command:

python lucidserver/server.py

By default, the API will run on http://127.0.0.1:5000/, and you can make requests to the API using tools like Postman or integrate it into your own applications.

Optionally, you can start the server locally with heroku, here is an example for running locally on heroku using a windows machine:

heroku local -f Procfile.windows

Then simply follow the steps above for using ngrok.

Heroku Deploy

Constraints

• Slug Size Limit: Slugs can't be larger than 500MB. If the slug is too large, consider using Docker or reducing the content included in the slug.
• Memory Limit: Memory cannot exceed 500MB for a single dyno. Monitor and optimize memory usage as needed.
• Dynamic Model Import: Models will be dynamically imported and used at runtime to keep the slug size manageable.
• Docker Deployment: If the slug is too large, consider deploying with a Dockerfile instead of a Procfile.

Tutorials

• Getting Started with Heroku
• Deploying with Docker on Heroku
• Heroku Slug Compiler

Commands

• heroku create (create a new Heroku app)
• heroku stack:set heroku-22 (set stack for Procfile deployment)
• heroku stack:set container (set stack for Dockerfile deployment)
• git push heroku main (push code to Heroku and start build)
• heroku logs --tail (view build and runtime logs)
• heroku config:set KEY=VALUE (set environment variables via Heroku CLI)
• heroku local -f Procfile.windows (run server locally on a windows machine)

Environment Variables

Must set ENV variables in Heroku dashboard or with Heroku CLI tool:

CLIENT_TYPE='POSTGRES'
POSTGRES_CONNECTION_STRING=YOUR_URI_CONNECTION_STRING