Closed jameslahm closed 8 months ago
Hello! I apologize for the inconsistencies, which stem from variations in the code versions. To enhance performance, consider adjusting prompts for specific classes. I am confident that there is substantial room for further improvement in MVTec through prompt tuning.
Thank you!
Thank you for your great work! I try to reproduce the results on MVTec-AD following the instructions in README, but the pixel F1 score on MVTec-AD is 37.65 rather than 39.40 which is reported in the paper. Do I need to tune some hyper-parameters to obtain the same result? Thank you!