-
Spike
-
Resolution: Unresolved
-
Undefined
-
None
-
None
-
False
-
-
False
-
-
- What are/will be the arguments of the APIs? What do we need to enable to ensure a complete InstructLab workflow?
- Users will need a UUID per request submitted - that can be sent as an (optional?) arg to the next job kicked off.
- Filesystem management - what happens to user artifacts:
- Input data, qna.yamls
- Document schema received after pre-processing
- Does user get SDG data? Trained model artifacts? In either case, where do these larger artifacts live?
- is related to
-
RHELAI-3619 Enable an end-to-end InstructLab workflow
-
- New
-