Open SofiaChar opened 7 months ago
Would be convenient to have the opportunity to pass the pipelines parameters when starting the pipeline.
For example we have this pipeline defined:
- pipeline: name: training parameters: - name: input-csv targets: - data-preprocessing.parameters.input-csv - name: embeddings-fr targets: - classification-model-training.parameters.embeddings-fr nodes: - name: fetch-dump-date type: execution step: fetch-dump-date - name: data-preprocessing type: execution step: data-preprocessing - name: classification-model-training type: execution step: classification-model-training edges: - [fetch-dump-date.metadata.formatted_dump_date, data-preprocessing.parameter.csv-dump-date] - [data-preprocessing.output.dataset-prod.pickle, classification-model-training.input.input-df]
When staring the pipeline pass the parameters:
vh pipeline run --adhoc training --input-csv=datum://018f3966-85d4-4e02-3fb9-bdb5e0c7bf75 --embeddings-fr=datum://018f399c-381d-eb61-730c-87e65c33994b
Would be convenient to have the opportunity to pass the pipelines parameters when starting the pipeline.
For example we have this pipeline defined:
When staring the pipeline pass the parameters:
vh pipeline run --adhoc training --input-csv=datum://018f3966-85d4-4e02-3fb9-bdb5e0c7bf75 --embeddings-fr=datum://018f399c-381d-eb61-730c-87e65c33994b