Config separate from weights on the Build Your Own AI Platform track. Thresholds, prompts, and feature flags change behavior without a new train. Version them. A prompt change is a release.
This lesson assumes you already worked through Promotion is a release.
The idea in practice
Store config beside the model version it was approved with. Log the config id on each request.
A concrete check
goal = {
'track': 'Build Your Own AI Platform',
'lesson': 'Config separate from weights',
}
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
Editing a prompt in production with no record. 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 three settings that are config, not weights.
Self-check
- Say Config separate from weights 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.