aghabidareh/NNs-from-scratch ? reverse-engineered prompt

Reverse engineered prompt

Build me a simple handwritten digit recognizer in a notebook, using only basic Python and math, no NumPy or other machine learning libraries.

I want it to train on MNIST from scratch and show the full process clearly, including loading the data, preparing it, building the neural network by hand, training it, and checking how well it works on test images. Please keep it easy to follow, like a learning project, with comments and explanations in the notebook so I can understand what each step is doing.

It would be great if it can also save a few example outputs or results in the outputs folder, and use the data folder for the dataset. If you need to look up current details for MNIST or notebook best practices, go ahead and do that.

Are you gonna build this?

make sure you review the code using coderabbit

Try freeSponsored — opens CodeRabbit in a new tab