scripts/run_pipeline.sh

#!/usr/bin/env bash
set -euo pipefail

SAMPLES="${1:-1000000}"
WORKERS="${2:-8}"

echo "==> Generating ${SAMPLES} labeled SQL samples..."
python -m src.generate_dataset --samples "${SAMPLES}" --workers "${WORKERS}"

echo "==> Training classifier..."
python -m src.train

echo "==> Evaluating..."
python -m src.evaluate

echo "==> Done. Model at models/sql_error_classifier.joblib"