ybracke / transnormer

A lexical normalizer for historical spelling variants using a transformer architecture.
GNU General Public License v3.0
6 stars 1 forks source link

Dev change save and logging strategy #97

Closed ybracke closed 1 month ago

ybracke commented 1 month ago

So far, the Trainer has saved, evaluated and logged according to X/Y/Z steps. However, I actually wanted to save once per epoch and evaluate/log n times per epoch. Now I have changed this to: 1) saving after X epochs, 2) logging/evaluating after every 0.5 epoch. A CustomCallback is used for this.