search

brjathu / iTAML

Official implementation of "iTAML : An Incremental Task-Agnostic Meta-learning Approach". CVPR 2020
96 stars 16 forks source link

about adding classes #9

Closed zhukaii closed 4 years ago

zhukaii commented 4 years ago

The function 'add_classes' in basic_net.py is not called in the code, so is the number of final classification nodes(10 or 100) of the network fixed? Is there any other place in the code where the number of nodes(classifier) increases？Thanks!

brjathu commented 4 years ago

Hi @zhukaii , yes we do not add new neurons to the model in the process, we simply had 100 neurons at the last layer, and train the model, however, this is same as adding new neurons, since the gradients are calculated for each task separately. You can still call add_classes, functions, it would be the same!

zhukaii commented 4 years ago

Thank you！