You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We will likely use multiple versions of Docker images. To avoid confusion, we should use version numbers (probably more intuitive than, e.g., hash IDs) rather than a fixed name with varying images under that name.
TODO
think about requirements
think about implementation
implement
The text was updated successfully, but these errors were encountered:
Following #22, the image would contain the code for local training. We cannot easily use that code also for the swarm training (I wouldn't know how), but we could check before starting the admin console that the code version to be sent out by NVFlare matches the version contained in the image to avoid accidentally using non-matching versions. Do we want that?
We will likely use multiple versions of Docker images. To avoid confusion, we should use version numbers (probably more intuitive than, e.g., hash IDs) rather than a fixed name with varying images under that name.
TODO
The text was updated successfully, but these errors were encountered: