optimizers · meta-research

Prodigy

Parameter-free optimiser that estimates the learning rate during training rather than taking it as a hyperparameter, building on the D-Adaptation line with a step-size estimate that adapts faster. Removes the one hyperparameter most often re-tuned when model size, batch size or schedule changes.

Verify

What this registry has checked

Quarantine. We hold the publisher’s own published hash for these weights and have recorded it against this entry. We have not downloaded and hashed the weights ourselves, so the hash is the publisher’s claim, not our measurement.
Re-check this yourself → Source → Full record →

What is recorded

Publishermeta-research
OriginUnited States
Licenceresearch-paper

Elsewhere in the registry

All optimizers → Browse with filters → What will run on my hardware? →

Registry record as of 2026-08-14. Identity, licence and parameter count are properties of a release and do not move; tier can, and the verify link above re-checks it against the live registry rather than this page.