Competition and Development resources¶

Honorary mention to Huggingface, a platform where you can store an infinite amount of models and data

Neural Network Development resources¶

Resources here will be model based, Ideally take a look at the tutorials before coming here

The most common location for every competition, Kaggle

Here's a couple designed to get you learning the format and maybe explore new kinds of NNs and data science you haven't seen before:

MNIST Data digit recognization, basically the very first useable NN anyone makes, but you can also explore the code and discussions to figure out how people optimize every aspect of a NN
Titanic competition, basically the main intro to kaggle
Housing prices, more around data science and non NN methods, but will definitely teach you that a NN is not suitable for every situation, and there are much better and lighter options out there
Space titanic, Titanic 2, try it after you're done with the first one

Still learner based competitions, but these are either meant for learning specific NNs or specific hardware

LLM classification, a starting point for anyone wanting to classify complex text
I’m Something of a Painter Myself, If you're interested in AI generated imagery of any kind, this is your starting point
Petals to the Metal, an intro to TPU chips, which google has in massive supply and provides large amounts to for student projects (look up google TRC), good to know how to use their hardware for when you need to scale up a project