Approve Models in the Model Registry
Approve models in the Model Registry with the ModelApproval.update method, which creates a ModelApprovalService.Record entry to track approvals.
Python
approval_metadata = c3.ModelApproval.Update.Status(modelUri = entry.uri,
modelVersion = entry.ver,
scope = "Approval Scope",
status = "Approved")
c3.ModelApproval.update(approval_metadata)Files can optionally be attached to a ModelApproval.Status to provide supporting evidence for the approval.
Python
file = c3.File.fromString("ModelApproval/metrics_file.txt").writeString("test")
files = {"test_file": [file.url]}
approval_metadata = c3.ModelApproval.Update.Status(modelUri = entry.uri,
modelVersion = entry.ver,
scope = "Approval Scope",
status = "Approved")
c3.ModelApproval.updateWithFiles(approval_metadata, files)Note on using scope of ModelApproval.Status
Scope is used to differentiate for what uses the model is approved. For example, an organization may want to only approve a model for a use in certain applications.