the function of stats() in train_classifier_fed.py

diaoenmao / HeteroFL-Computation-and-Communication-Efficient-Federated-Learning-for-Heterogeneous-Clients

[ICLR 2021] HeteroFL: Computation and Communication Efficient Federated Learning for Heterogeneous Clients

MIT License

150 stars 33 forks source link

the function of stats() in train_classifier_fed.py #11

Closed ghost closed 1 year ago

ghost commented 1 year ago

Hi, thank you for the excellent work. I have a little question, what is the function 'stats' used for? There is a for loop inside to feed data to the testmodel, however, I don't find the optimizer to update the model and I assume it just "feed" without anything else, then why does it exist. However, when I comment the for loop, the global test accuracy drops, it's weird to me. I appreciate if you can explain it!

diaoenmao commented 1 year ago

The stats() is used for computing the statistics used in static batch normalization.