fine-tune-distilbert-on-jsonl-with-manual-encoding

Generates a Python script to fine-tune a DistilBert model on a JSONL dataset containing 'question' and 'answer' columns. The script uses manual label mapping (avoiding sklearn), includes progress logging, error handling, and model evaluation.