Part 3 from PR #25 .
An ability to specify whether to use the "inference" backend as default or the "learning" backend as default. This would be a config option on the client and broker.
- Simplest case: This config would apply plugin wide. An environment variable and/or parameter to the client.
Above-and-beyond-case (absolutely not necessary, attempt if only feeling very motivated :)
The config applies per session. The plugin would intelligently track what engines have been used, and apply the default to that session's (and digital twin's) generated tasks. This would likely require quite a bit of restructuring.
Part 3 from PR #25 .
An ability to specify whether to use the "inference" backend as default or the "learning" backend as default. This would be a config option on the client and broker.