Open Zesheng666 opened 1 year ago
I've tried the DN_DAB model several times. After 12 epochs, the AP is only 0.299, and after 34 epochs, it's just 0.366. These results are far from what is stated in the paper. Why is this the case?
Have you modified something in the code or in the running command?
All the config files remain as specified in the original code, except for the dataset addresses.
I've tried the DN_DAB model several times. After 12 epochs, the AP is only 0.299, and after 34 epochs, it's just 0.366. These results are far from what is stated in the paper. Why is this the case?