Fine-tuning on the AI Learning track. Fine-tuning continues training on your examples. You must keep a validation set the fine-tune does not see. You also inherit the base model's licenses, biases, and context limits.
This lesson assumes you already worked through Transfer from another task.
The idea in practice
Use a small learning rate. Watch for catastrophic forgetting of the behavior you still need. Save the base checkpoint so you can roll back.
A concrete check
goal = {
'track': 'AI Learning',
'lesson': 'Fine-tuning',
}
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
Fine-tuning on the test set, or on data you are not allowed to store. 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 what you would log so you can prove what data the fine-tune saw.
Self-check
- Say Fine-tuning 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.