Me i will send you a big requirements doc. you shall proceed to send me a file tree. I will, ask you for a few files at a time. You shall give me the code step by step. deal?

ChatGPT Understood, you want to send a requirements document and want a file tree in response. Will provide code files step-by-step upon request. Deal accepted. Note: I can't literally send files, but I can provide text-based representations of what the code files should contain.

we are using fastapi.

Services

Gateway

• This gateway api is the main point of contact for the interfaces
• It has the following main components;
  • a basic server
  • an endpoint to receive a request to create a memeRequest.
  • a util folder that incl:
    • a service file called ai_service that handles tasks executed by AI’s
      • initializes an openAI gpt3 model.
      • includes a function called call(string, maxTokens=1000), temperature set to 0.5
      • includes a function called generateCaptions(memeRequest, template) uses a template string stored under prompt_templates/get_captions.txt
        • calls client.call(templated prompt).
        • splits it into ‘template.line’ lines
    • a file that handles common s3 functions like initializing connections, getting stuff, putting stuff
  • a logic module that includes:
    • memeService
      • this will contact an external memeClient
      • a function getMatchingMemeTemplates(memeReq, n=6) that takes in a MemeRequest and returns a list of n Templates
        • for each template, asyncly calls generateMeme(memeReq, template_i)
      • a function called generateMeme(memeReq, template)
        • template DataType includes a property called ‘Description’.
        • calls the ai_service.generateCaptions(memeRequest, template)
        • calls the memeClient.generateMeme(templateId, captions)
        • using storageModule, stores the meme (in s3) as follows
          • {memeRequest.companyName}/{template_id}_{uuid}
  • a storage module that includes:
    • a file with all the models
      • the data type MemeRequest is:
        • companyName: string
        • productDescription: string
      • the data type Template is:
        • lines: number
        • description: str
        • id: str
    • a file with all the functions to store/ retrieve the models ChatGPT Based on your FastAPI project requirements, here is a proposed file tree: bash

├── main.py  # Entry point
├── services
│   └── gateway
│       ├── server.py  # Basic server setup
│       ├── endpoint.py  # API endpoints
│       └── util
│           ├── ai_service.py  # AI-related tasks
│           └── s3_common.py  # S3 utilities
│       ├── logic
│           └── meme_service.py  # Meme logic
│       └── storage
│           ├── models.py  # Data models
│           └── storage_functions.py  # Functions to interact with storage
└── prompt_templates
    └── get_captions.txt  # Template string for caption generation

main.py: FastAPI app initialization, imports modules, and routes.

services/gateway/server.py: Basic FastAPI server setup with initial routes and middleware.

services/gateway/endpoint.py: Endpoint to handle meme requests.

services/gateway/util/ai_service.py: Initialization of OpenAI GPT-3, call() and generateCaptions() functions.

services/gateway/util/s3_common.py: S3-related utility functions like initializing connections, get, and put operations.

services/gateway/logic/meme_service.py: Includes getMatchingMemeTemplates() and generateMeme() functions.

services/gateway/storage/models.py: Defines MemeRequest and Template data types.

services/gateway/storage/storage_functions.py: Functions for storing and retrieving models.

prompt_templates/get_captions.txt: Template string for AI caption generation.

Let me know which files you would like to explore first.