This repository contains our implementations for "Solving Delayed Feedback by Dual-level Unlearning in Online Continuous Conversion Rate Prediction".
- numpy==1.23.5
- pandas==1.5.3
- python==3.10.9
- scipy==1.10.0
- torch==2.0.0
- pyyaml==6.0
- tqdm==4.65.0
-
Download the dataset under './data/'
-
Run the following command to pretrain
python main.py --mode pretrain --method Pretrain-
Run the following command to train and evaluate the model
python main.py --mode stream --method Oracle