Skip to content

Model Training & Adaptation

Early stopping

Stopping AI training when checks show it is no longer improving.

Example

Training stops after independent checking performance fails to improve for several checks.

Why people use it

It avoids spending more training time after useful performance stops improving.

What you'll hear

“Stop when it stops getting better on fresh examples.”

What this means for you

Track a suitable independent checking measure rather than optimizing training fit alone.

Can you control it?

Developer-only

The people building or running the AI choose this setup. An everyday user generally needs their help to change how this part works.

Common questions

Does one worse check always stop training?
No. Many setups allow a period without improvement before deciding to stop.
Can stopping too early hurt performance?
Yes. The system may not yet have learned the useful patterns it could learn.
Why not judge only on the examples used for learning?
It can keep improving on familiar examples while becoming less useful on new ones.

Related terms

Still have questions?

Up to 500 characters.

Ask LATHIC about AI. Relevant glossary entries may be included.

Your question, the glossary entries it matches, and a rotating pseudonymous identifier go to Microsoft Azure’s OpenAI service through Vercel AI Gateway to generate an answer. Zero retention and no training are required of the provider, and LATHIC does not save your question or answer. Privacy Notice