Open Models Almanac
Methodsoptimizationlearning-rate schedule

single mention · optimization

Learning-rate decay

The learning rate is reduced during fine-tuning from an initial value to a lower value after a specified number of gradient steps.

The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See learning-rate schedule for the methods filed beside it.

Also called Learning-rate decay during fine-tuning.

Filed under optimization › learning-rate schedule