CFT

Train

bash run.sh

Inference

bash infer.sh

You should use your own model and data path.

Evaluate

The original grounding method(BIGRec):

python evaluate_acc.py

The adjusted grounding method:

python eval_acc_2.py

To get the distribution of item popularity, you can run the following code:

python evaluate_all_pop.py