[Bug]: Missing Metric Configuration in Sweep Example

[GabrieleDian]

Tell us what happened?

Location

• File path:colabs/boosting/Credit_Scorecards_with_XGBoost_and_W&B.ipynb
• Section: HyperParameter Sweep

The current sweep example miss the metric configuration:

sweep_config = {
    "method": "random",
    "parameters": {
        "learning_rate": {
            "min": 0.001,
            "max": 1.0
        },
        "gamma": {
            "min": 0.001,
            "max": 1.0
        },
        "min_child_weight": {
            "min": 1,
            "max": 150
        },
        "early_stopping_rounds": {
            "values": [10, 20, 30, 40]
        },
    }
}

As a result, Sweep visualizations show parameter values (e.g., min_child_weight) as the optimization goal instead of the actual metric

Suggested Fix

Add the missing metric configuration:

sweep_config = {
    "method": "random",
    "metric": {
        "goal": "minimize",  
        "name": "val_log_loss"  
    },
    "parameters": {
        # ... existing parameters
    }
}