Writing it versus using a library on the How to Make AI Models track. Use a library for training loops and layers unless you are studying the mechanic. Write a tiny version once so you know what the library is doing.
This lesson assumes you already worked through Small models on purpose.
The idea in practice
A 20-line linear model you can read is worth more than a framework you cannot debug.
A concrete check
goal = {
'track': 'How to Make AI Models',
'lesson': 'Writing it versus using a library',
}
checks = [
'input available at decision time',
'score matches the real decision',
'failure case written down',
]
print(goal['lesson'])
for item in checks:
print('-', item)
Run the sketch locally if you have Python. The printout is a reminder of the checks, not a trained model. Replace the strings with the real inputs from your own example before you treat it as a design.
What usually goes wrong
A custom autograd you do not test against a finite difference. When this happens, stop adding parameters or tools. Fix the check, the data, or the permission, then run the same example again.
What to write down
- The input you are allowed to use at decision time.
- The output and the score or pass rule.
- One failure you will test on purpose.
- What you will not claim the system can do.
Practice
Name one piece you would implement by hand for learning, and one you would import.
Self-check
- Say Writing it versus using a library in one sentence that mentions an input and an output.
- Name the failure mode in this lesson and the check that would catch it.
Done when: you can explain this lesson without the page open, and you have a written failure case.