|
|
| """
|
| Quick Data Analysis Runner
|
| Runs the comprehensive data analysis and provides immediate recommendations.
|
| """
|
|
|
| import sys
|
| import argparse
|
| from pathlib import Path
|
|
|
|
|
| sys.path.insert(0, str(Path(__file__).parent))
|
|
|
| from analyze_training_data import TrainingDataAnalyzer
|
|
|
| def main():
|
| """Run data analysis and provide recommendations"""
|
| parser = argparse.ArgumentParser(description="Quick CVE Training Data Analysis")
|
| parser.add_argument("--dataset", default="data/training_datasets/enhanced_training_dataset_with_mitigations.json",
|
| help="Path to training dataset")
|
| parser.add_argument("--sample-size", type=int, default=200,
|
| help="Number of samples to analyze")
|
| parser.add_argument("--no-viz", action="store_true",
|
| help="Skip visualization generation")
|
|
|
| args = parser.parse_args()
|
|
|
| print("🔍 CVE Training Data Analysis")
|
| print("=" * 50)
|
|
|
|
|
| dataset_path = args.dataset
|
| if not Path(dataset_path).exists():
|
| print(f"❌ Dataset not found: {dataset_path}")
|
| print("Please run dataset_preparation.py first to create the training dataset")
|
| return False
|
|
|
|
|
| analyzer = TrainingDataAnalyzer(dataset_path)
|
|
|
| print("📊 Running comprehensive data analysis...")
|
| print("This will analyze:")
|
| print(" - Instruction diversity")
|
| print(" - Input diversity")
|
| print(" - Output patterns")
|
| print(" - Duplicate detection")
|
| print(" - Data distribution")
|
| print()
|
|
|
| success = analyzer.run_full_analysis(sample_size=args.sample_size, generate_viz=not args.no_viz)
|
|
|
| if success:
|
| print("\n📁 Analysis results saved to 'data_analysis/' directory")
|
| print("📊 Check 'data_analysis_visualization.png' for visualizations")
|
| print("📝 Check 'data_analysis_report.md' for detailed report")
|
|
|
|
|
| red_flag_count = sum(1 for results in analyzer.analysis_results.values()
|
| if 'red_flag' in results and results['red_flag'])
|
|
|
| print("\n🎯 IMMEDIATE RECOMMENDATIONS:")
|
| print("=" * 50)
|
|
|
| if red_flag_count > 2:
|
| print("❌ CRITICAL: Multiple data quality issues detected!")
|
| print(" → Fix data quality before training")
|
| print(" → Consider data augmentation")
|
| print(" → Remove duplicate examples")
|
| print(" → Increase validation set size")
|
| elif red_flag_count > 0:
|
| print("⚠️ WARNING: Some data quality issues detected")
|
| print(" → Use overfitting fixes in training pipeline")
|
| print(" → Start with small subsets (--subset 10)")
|
| print(" → Monitor training closely")
|
| print(" → Consider data cleaning")
|
| else:
|
| print("✅ GOOD: Data quality looks acceptable")
|
| print(" → Safe to proceed with training")
|
| print(" → Use standard training settings")
|
| print(" → Monitor for overfitting during training")
|
|
|
| print("\n🚀 NEXT STEPS:")
|
| print("1. Review the detailed report in data_analysis_report.md")
|
| print("2. Run training test: python test_overfitting_fixes.py")
|
| print("3. Start training: python src/training/fine_tuning_pipeline.py --subset 10")
|
|
|
| return True
|
| else:
|
| print("❌ Data analysis failed!")
|
| return False
|
|
|
| if __name__ == "__main__":
|
| success = main()
|
| sys.exit(0 if success else 1) |
