Inspiration
Making AI models is hard, and even harder when you don't know how to use machine learning libraries. Even though these libraries are meant to make the process easier, sometimes its just confusing and overall annoying trying to make a machine learning model. This can take days, to even weeks to perfect, and it may take a large chunk of time out of a side project that doesn't have the model as its main priority.
What it does
pAIMLess is a lightweight application that allows users to quickly create and deploy neural networks. You basically give it a csv to train and tweak some settings of the machine learning model, and the program automatically trains the model. You can also host it on a server handcoded by us, and send requests to it.
How we built it
We used Tensorflow with Python, processed data with pandas, created a httpserver, and the interface is with tkinter.
Challenges we ran into
We had many, many Git merge errors. Not everyone in our group knew much about neural networks, so they had to learn a bit. However, by the end, it turned out fine as we pulled through.
Accomplishments that we're proud of
We were able to make a working prototype of a no-code neural network builder that seemed daunting at first. It is actually really cool how you can send requests to the api/server that we hosted the model on.
What we learned
We learned skills for collaboration on Git, and more about AI in general. We also learned about csv reading and httpsserver making.
What's next for pAIMLess
Adding more customization (more types of layers), more templates, and sharing templates.
Log in or sign up for Devpost to join the conversation.