Machine Learning
ML Pipeline Skill
Skill for working with the ML extraction pipeline — spaCy NER model training, regex-based attribute extraction, production extractor that combines both methods, resolver logic for cross-validation between ML and regex, and model evaluation. Use this skill whenever the user mentions NER, spaCy, extraction accuracy, mileage/fuel/power/year extraction, labeling data, training or retraining the model, resolvers, ProductionExtractor, error analysis, F1 score, training reports, or anything in the ml/ directory. Also trigger for labeling/ directory work (label_data_assisted.py, training data preparation).