Enhanced Evaluation Metrics with Macro/Micro Scores and Per-Label Analysis

[Xiaomin-HUANG]

Added Macro F1 and per-label metrics to complement the existing Micro F1 score, enabling better visibility into model performance across different entity types.

What's New

• Macro F1: Unweighted average across all labels (0.51)
• Per-label breakdown: Precision, Recall, and F1 for each entity type
• Formatted table: Sorted output for quick identification of best/worst performers

Enhenced output : {
        
        "per_class":{"tag1":{"precision":float, "recall":float,"f_score":float},
                    "tag2":{}...
                    },
        "micro":{"precision":float, "recall":float,"f_score":float},
        "macro":{"precision":float, "recall":float,"f_score":float},
        }

Formatted table eg :