Creating a new Trainer

To create a new DataDreamer trainer class to support a new training library or training technique, you will want to subclass the Trainer class. You can see example implementations of various trainers by clicking on the [source] links on the Trainers page. These may be helpful as reference implementations.

Contributing

If you would like to contribute the new trainer class you created to DataDreamer for others to use, see the Contributing page. If applicable, please ensure your implementation includes appropriate metadata, such as a link to the model card of the model being trained, the model’s license, and the model and training technique’s citation information.