search

theislab / cpa

The Compositional Perturbation Autoencoder (CPA) is a deep generative framework to learn effects of perturbations at the single-cell level. CPA performs OOD predictions of unseen combinations of drugs, learns interpretable embeddings, estimates dose-response curves, and provides uncertainty estimates.
BSD 3-Clause "New" or "Revised" License
83 stars 17 forks source link

v0.5 "ValueError: Expected a parent" in tutorial #20

Closed AxKo closed 1 year ago

AxKo commented 1 year ago

Hello Everybody,

I tried to follow the SciPlex2 tutorial with the new version 0.5. After adjusting ae_hparams and trainer_params I call model.train() and the CPA crashes with "ValueError: Expected a parent".

Some debugging shows that the problem is in _validate_callbacks_list() when the is_overridden() method is called on the EarlyStopping callback. Since is_overridden() is called with 'parent=None' and there is no code to set 'parent' for an EarlyStopping instance the ValueError is raised.

What am I doing wrong?

Any help is appreciated, Axel

znavidi commented 1 year ago

I have the same issue! Would appreciate any guidance.