Hello, author. First of all, thank you for your work. While reproducing your work, I found that the contrastive loss is difficult to converge below 1. Is this normal? What value did it converge to during your training? I am using the CAT-NET dataset. I would greatly appreciate it if you could help clarify this for me.
Hello, author. First of all, thank you for your work. While reproducing your work, I found that the contrastive loss is difficult to converge below 1. Is this normal? What value did it converge to during your training? I am using the CAT-NET dataset. I would greatly appreciate it if you could help clarify this for me.