Start a new agentic ML training pipeline run. This initiates the full ML workflow: data preparation -> label selection -> feature engineering -> model training -> deployment -> reporting -> monitoring. The pipeline will pause at each phase listed in require_approval and wait for a decision via submit_decision().
If True, automatically proceed through phases without approval
require_approvallistdefault: None
List of phases requiring human approval. Defaults to the server's gate configuration. Valid phases: label_selection, data_preparation, feature_engineering, model_training, model_deployment, report_creation, monitoring_creation
Check if the pipeline is waiting for a human decision. When the pipeline reaches a phase listed in require_approval, it pauses and creates a pending decision. Call this to see what decision is needed, then use submit_decision() to provide your choice.
Submit a decision for a pending approval in the pipeline. The required fields depend on the decision_type from get_pending_decision(). Common patterns: - Label selection: decision_type="label_selection", choice_index=N - Feature engineering: decision_type="feature_engineering", apply_indices=[...], skip_indices=[...] - Model training: decision_type="model_training", choice_index=N - Deployment: decision_type="model_deployment", action="approve" or "skip" - Report: decision_type="report_creation", report_config={...} - Monitoring: decision_type="monitoring_creation", monitoring_config={...}
Type of decision (matches pending decision's decision_type)
choice_indexintdefault: None
Index of chosen option from the options list
actionstrdefault: None
Action string (e.g., "approve", "skip", "reject")
custom_valueanydefault: None
Custom value for decisions that accept free-form input
label_typestrdefault: None
Label type for label selection decisions
apply_indiceslistdefault: None
Indices of items to apply (e.g., feature engineering steps)
skip_indiceslistdefault: None
Indices of items to skip
selected_indiceslistdefault: None
Indices of selected items
selected_optionslistdefault: None
List of selected option values
selected_featureslistdefault: None
List of selected features
monotonic_constraintsdictdefault: None
Feature -> direction ("increasing"/"decreasing") map for training approvals. Passing a dict — even an empty one — marks the constraint proposals as reviewed; pass {} to train unconstrained after reviewing, or omit (None) to leave the server default behaviour.
Send a chat message to the agentic pipeline and get a response. Use this to ask questions about the current state of the run, request explanations, or provide additional context.
Get the preprocessing DAG (directed acyclic graph) visualisation data for a run. Shows how preprocessing steps are connected and their order of execution.
