About reproduce

[LFM-bot]

Hi, I tried to reproduce your work using the provided 1% poisoned data. However, the results I got are about half of the report, which makes me confused. Can you tell me some details about training the victim models (CL4SRec and S3Rec)? Such as the data augmentation methods for SSL pretraining (CL4SRec usually uses crop, mask, and reorder operations, while S3Rec uses mask item prediction and mask segment prediction), the augmentation ratios, the epochs of pretraining and fine-tuning, and the early stopping patience (as I noticed the hit-ratio of the attack item is decreasing as the fine-tuning goes). Also, I would appreciate it if you could provide the code for victim model training.