roadmap v0.1.0

[hsinatfootprintai]

virtuscale provides native integration to the underlying backend (e.g. kfp). This doc lists feature requests available.

• training

• support TensorFlow

• support PyTorch

• support multi-step training architecture

• add training metrics with push model

• support model registry

• model storage

• support pvc & s3

• support versioning

• support copy and tags (like dockerimages)

• inferencing

• support kserve

• support swagger

• support inferencing metrics