Ask HN: Would an open source SDK for federated/distributed ML be useful?
2 comments
A little bit off topic, we were actually thinking about implementing federated learning for a client but in the end didn't went through with it, I've just read about it in a research environment.
What are your use cases for it, are you using it commercially or just for research?
What are your use cases for it, are you using it commercially or just for research?
Interesting! Well I haven't really thought about its implications commercially, what would the benefits for be really? I was considering it more from the research perspective, but i'll give the commercial applicability some thought.
Also, so you think there are any advantages to automating and distributing ML operations for large companies/orgs?