Getting the trained model's predictions

[The-BlackCoat]

Hello. Thank you for this helpful code repository.

I have a question regarding using a trained model and obtaining its predictions.

In the documentation, you mention that:

"""

Testing the trained model

trainer.test(test_dataloaders=dataset.test_dataloaders())
"""

However, executing this code provides us only with some error metrics (e.g., test loss, accuracy), and not the actual trained model's predictions.

Can you help me with getting the model predictions and then convert the predictions to a NumPy array?
I need the model's predictions to calculate some statistics on them (for a classification task), that's why I need them in a NumPy array.

Thank you.
Best regards

[rkalahasty]

Hey, was wondering if you figured out how to get the model predictions? I am also doing a classification task