A Component taking dict[str, Input[Artifacts] as input in kfp pipeline is not even starting

Question

from typing import Dict
from kfp.dsl import component, Input, Artifact

@component
def best_model(model_metrics_jsons: Dict[str, Input[Artifact]]):
    import os
    
    for model_name, metrics_input in model_metrics_jsons.items():
        metrics_json_path = metrics_input.path
        print(f"Model: {model_name}")
        print(f"Metrics JSON Path: {metrics_json_path}")
        
        if os.path.exists(metrics_json_path):
            with open(metrics_json_path, 'r') as file:
                metrics_json = file.read()
                print(f"Metrics JSON Content: {metrics_json}")
        else:
            print(f"Metrics JSON Path does not exist: {metrics_json_path}")

from kfp import dsl

@dsl.pipeline(name='dynamic-datasets-pipeline')
def dynamic_datasets_pipeline():
    # Example components
    preprocess_task = preprocess_data()  # Your preprocessing component
    evaluate_model_tasks = evaluate_models()  # Your evaluation tasks

    # Create the dictionary of model metrics
    model_metrics_dict: Dict[str, Input[Artifact]] = {
        selected_model_names[i]: evaluate_model_tasks[i].outputs['model_metrics_json']
        for i in range(len(selected_model_names))
    }
    
    # Call the best_model component with the dictionary
    best_model_task = best_model(model_metrics_jsons=model_metrics_dict)

I suspect there may be an issue with how the dictionary model_metrics_dict is being passed to the best_model component or how the components are connected. Can someone help me identify the issue or suggest a solution?