ylsung / gnn_few_shot_cifar100 Goto Github PK

samples = random.sample(self.data[class_], num_shots+1)
I'm not sure if I understand it right, can you explain it? thank you very much!
best wishes!

I am not able to find out how class overlap between train and test data is being maintained. Typically in few shot learning evaluation, few classes are never shown to the model, and on test time we evaluate the model by giving it N examples from those classes that have not been shown to the model. I am not sure if this separation is maintained in the code.