New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Create TrainingPhase class and/or add "add_training_phase" method to ParametricDR #1

Closed

einbandi opened this issue Jul 5, 2022 · 1 comment

Assignees

Projects

Implement base functionality

Collaborator

einbandi commented Jul 5, 2022

This class/method should allow users to set all the parameters required for a training phase. These are:

number of epochs
sampling method
batch-wise relations (specified as a dict, so custom loss functions can acces them)
loss function (which gets passed everything [model, global relations, batch-wise relations, batch] and accesses parts through keys)

einbandi self-assigned this

Collaborator Author

einbandi commented Jul 5, 2022 •

edited

There are three possibilities for the API:

specify phases at training time: dr.train(phases=<list of TrainingPhase objects>)
specify phases in __init__: `dr = ParametricDR(..., training_phases=)
add phases one by one before training dr.add_training_phase(**kwargs)

Choose one or allow multiple?

einbandi added this to To do in Implement base functionality

einbandi moved this from To do to In progress in Implement base functionality

einbandi closed this as completed in

c85d43a

einbandi moved this from In progress to Done in Implement base functionality

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment