/names-ai

This project is a fully deployed containerized ML model served via REST API. The model generates unique fictional names, powered by MLP-based deep learning.

Primary LanguagePython

This project is a fully deployed containerized ML model served via REST API.

The character-level model generates fictional names, powered by a 6-layer MLP deep learning neural network. Create your next character's name with AI.

How to use

This microservice exposes 2 REST API endpoints:

  1. {your_server}/
  2. {your_server}/names/{number}?seed={seed_setting}

The 1st endpoint returns a welcome message. The 2nd endpoint takes a required path parameter {number}, which is any positive integer, that specifies the number of generated names to return in the response body. The optional query parameter ?seed={seed_setting} can be used to control the source of randomness during name sampling for deterministic, reproducible results. seed_setting can be any positive integer. If the query parameter is omitted during the call, seed_setting defaults to 1 in the application.

What does it do

Each call to the {your_server}/names/{number} endpoint returns a list of new unique names generated by the character-level neural network. These names can be used in any creative process, such as fictional character name creation for novels, comic books, board games, movies and generative AI prompts.

For example, the following GET call: {your_server}/names/15?seed=99 will return the following response: {"names":["mariel.","avie.","drely.","revilli.","jahompson.","ayshanvitte.","sulaith.","joanna.","conriq.","mir.","yarie.", "legria.","zaiah.","brodeixrenley.","gia."]}

Note: some names generated might be close to or exact matches with names used today.