Searching hyperparameters on the How to Train Models track. Search learning rate, regularization, and model size on validation only. Budget the number of trials. Random search beats a huge grid when many knobs do not matter.
This lesson assumes you already worked through Training when data is small.
The idea in practice
Lock the search space before you look. Record every trial, including failures.
A concrete check
goal = {
'track': 'How to Train Models',
'lesson': 'Searching hyperparameters',
}
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
Hundreds of trials on the test set. 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
List three knobs you would search and one you would freeze.
Self-check
- Say Searching hyperparameters 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.