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
The skill oracle is currently implemented with GET and PUT API calls but has no mechanism for automatically ingesting more skill candidates when needed. I've purposely left logic for this out of the service (although you can peek at _push_once unittest to see how it might be partially implemented).
For the end to end skills-labeller we need some kind of component that:
Detects number of unlabeled candidates
When number of unlabeled candidates is low, pushes unlabeled candidates from the database/db into the skill-oracle (as unlabeled candidates, via PUT)
(discussion in #mlops will be held, see what's a simple way to do this)
The text was updated successfully, but these errors were encountered:
The
skill oracle
is currently implemented with GET and PUT API calls but has no mechanism for automatically ingesting more skill candidates when needed. I've purposely left logic for this out of the service (although you can peek at_push_once
unittest to see how it might be partially implemented).For the end to end skills-labeller we need some kind of component that:
db
into theskill-oracle
(as unlabeled candidates, via PUT)(discussion in #mlops will be held, see what's a simple way to do this)
The text was updated successfully, but these errors were encountered: