federatedai / fate-client Goto Github PK

Support fast run single component when backend execution is standalone

pipeline:2.0 support saving and loading pipeline without pickle

Is there any operation like flow table list -ns xxxxxxx to get table names

ml_metadata lib is not stable, it's better to use sqlite instead of ml_metadata

Standalone execturo's job and task status should be corrected to align with fate_flow executor

Does flow client or pipeline support dryrun mode?

The flow client or pipeline submit the job, and the flow server only parse the dag without running them.

It's only for testing the dag or pipeline

1. Add task: support multiple tasks
2. add get_output_data api
3. task.hosts: support add common pamameters to all hosts