GMvandeVen / continual-learning

PyTorch implementation of various methods for continual learning (XdG, EWC, SI, LwF, FROMP, DGR, BI-R, ER, A-GEM, iCaRL, Generative Classifier) in three different scenarios.
MIT License
1.53k stars 310 forks source link

about kafc fisher infromation matrix #22

Open fmaaf opened 1 year ago

fmaaf commented 1 year ago

Hi, thanks for your great work! I met the following error when I turn _fisherkfac to True :

raise Exception(f"Layer {label} does not have phantom parameters") Exception: Layer fcLayer1 does not have phantom parameters

Cloud you please help me find what's wrong with this exception? Thanks!

GMvandeVen commented 1 year ago

Thanks for your interest in my code, and thank you for raising this error. I haven't fully tested the option --fisher-kfac yet in all possible configurations (that's why I haven't included it in the documentation yet). I'm afraid at the moment I don't have time to look into this, my apologies, but I hope to get back to it in the near future.