The API is currently on hiatus due to operational expenses of GPU infrastructure. The front-end is still fully functional. Project accomplishments (team of 3):

• We aggregated data from multiple sources to get a specialized multilingual dataset of versified text.
• We trained and evaluated from scratch a compact language model in PyTorch
• We served and deployed the model securely with FastAPI, testing the maximum number of concurrent users the model can handle.
• We built a Next.js front-end to interface with the API. With a minmalistic zen design, features like visual and sound effects that can be deactivated, and a exportable cards to showcase the results