SAML
SAML copied to clipboard
Question about number of meta train and meta test
Thanks for your great work and code release. I have a question about training process. In this why num_meta_train is 2 ? Totally we have 5 source domains during training in this code, only use 2 domains for meta train and 1 domain for meta test. Am I right ? I am a little confused. Why not use 4 source domain for meta train and 1 source domain for meta test ? thanks for your help.
Thanks for your great work and code release. I have a question about training process. In this why num_meta_train is 2 ? Totally we have 5 source domains during training in this code, only use 2 domains for meta train and 1 domain for meta test. Am I right ? I am a little confused. Why not use 4 source domain for meta train and 1 source domain for meta test ? thanks for your help.
Hello,how to train my own dataset using the network? I couldn't find any related methods in the readme, and I'm looking forward to your answer. Thank you very much!