Normalization layers on the How to Train Models track. Batch norm and layer norm rescale activations so training can use larger rates. Batch norm depends on batch statistics. At inference you must use the running averages, not a batch of one by accident.
This lesson assumes you already worked through Initialization.
The idea in practice
Confirm the framework is in eval mode when you measure validation. Train mode uses batch stats.
A concrete check
goal = {
'track': 'How to Train Models',
'lesson': 'Normalization layers',
}
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
Reporting validation accuracy while the model is still in train mode. 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
Say what changes between train mode and eval mode for batch norm.
Self-check
- Say Normalization layers 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.