recommenders-team/recommenders
Auf GitHub ansehen[ASK] Running evaluation on test set for DKN
Open
#1.807 geöffnet am 4. Aug. 2022
help wanted
Repository-Metriken
- Stars
- (17.706 Stars)
- PR-Merge-Metriken
- (Durchschn. Merge 6T 16h) (10 gemergte PRs in 30 T)
Beschreibung
Description
I've have been using DKN Deep Dive, but I wanted to add evaluation on the test data. When I run model.run_eval(test_file) I am getting an error. I believe this is because the test data has users that are not present in the training data.
I am using the small MIND dataset. But I split my training, validation and test data a little differently. Training is all of the small training data but the last day. Validation is the last day of the small training data Test is all of the small validation data.
Is this the expected behavior?