few-shot
few-shot copied to clipboard
Queries about the MAML implementation
In line 84 few_shot/maml.py
loss.backward(retain_graph=True)
I don't understand the usage of this sentence. I wonder it aims to calculate the gradient of task_val_loss towards the initial parameters, but it has not been used in the latter codes. Could you explain the meaning of this sentence? Thanks a lot!
I'm interested to know the answer to this question, too. Thank you!
I wonder if this line is not necessary as there are no updates on the weights after this line, but I might be wrong