Can I train locally only for research purposes?

[Anja-Hobby]

This probably is a contradicting question to the whole Idea of Federated learning. However, for a research project, we want to compare the privacy and accuracy of different federated machine learning models.

But I get this error with registering:

File "...\anaconda3\envs\fed3ML\lib\site-packages\requests\models.py", line 960, in raise_for_status
    raise HTTPError(http_error_msg, response=self)
requests.exceptions.HTTPError: 503 Server Error: Service Temporarily Unavailable for url: https://service.eu-de.apiconnect.ibmcloud.com/gws/apigateway/api/

And when I try to run it without registering, it will lead to problems in aggregator.py.

[Anja-Hobby]

I just found out there is a research branch! I will mark this as closed :)
https://github.com/IBM/federated-learning-lib/tree/main/experiment_manager