NutritioNow Website

A website to help you answer all your nutrition-related questions!

Comment

Inspiration

We were inspired to make this website so that everyone can have access to information about the importance of nutrition.

What it does

Our website takes in a prompt and generates an accurate answer for the user. Purpose is to remove the need to search up nutrition facts multiple times by having all the information in one place. It saves time by giving you a concise answer instead of you searching through multiple different sources.

How we built it

We built the backend using Django with Python and built the front end using HTML, CSS, and Javascript.

Challenges we ran into

We spent a lot of time figuring out how to train a small compact model with answers specific to nutrition. However, despite training it with over a hundred questions and responses and the endless amount of prompts that could be related to nutrition, we soon realized our model could not handle all questions provided by the user. We decided instead to utilize a larger and more general model to guarantee our model could answer all of the user's questions regarding nutrition.

Accomplishments that we're proud of

Despite the challenges we faced, we managed to make a website that provides users with accurate answers to any questions they may have about a wide range of nutritional information.

What we learned

We learned that training a good AI model is very difficult, and even then cannot guarantee accurate results.

What's next for NutritioNow

Currently, our website can only be run locally on each user’s computer due to the size of the model we are using. Because it's running locally, the speed depends on the specs of the user's device. In the future, we’d like to find a way to host the website on a server, so that everyone can have easy access to the website!

Notes on the project itself

  • Has been tested on a Macbook (M1 Pro), Macbook (M2 Air), Linux (Thinkpad T15), but not Windows
  • Must have a least 4GB of RAM
  • Depending on the computer, speed may vary, if this were hosted, the speed would be consistent and quicker
  • To compensate for the whole category of nutrition, our chatbot relies on LLM (Large Language Models)
Share this project:

Updates