Primary Goals:

• Integration training host and inference host with SMO/Non-RT RIC/Near-RT RIC

H Feature Scope / Achievements: 

• Diversify training data source for Training host
  • Obtaining training data from DME in Non-RT RIC 
  • Creating Feature groups with data source and feature information

• Kserve adapter

  • Deploy and manage AI models in Near-RT RIC/Non-RT RIC

  • Integrate Inference host with O-Cloud( RICDMS ) and Management Func of RIC. 

  Diversify training data source for Training host
  • Obtaining training data from DME in Non-RT RIC 
  • Obtaining training data from influxDB in SMO

• Training pipeline Enhancement

  • Provide sample pipelines by default

  • Simplify the process of adding new training pipeline 

• AIMLFW feature enhancements
  • Options for edit, retrain and delete training jobs 

  Integrated installation

  • Provide install script of inference host in Near-RT RIC/Non-RT RIC 
  • Provide install script of training host in SMO

Sprint Demo