configs/full_with_eval.yaml

# Azure OpenAI configuration with credentials in the file
dataset:
   root_path: kdd-benchmark/input

agent:
  model: gpt-4.1
  api_base: https://api.openai.com/v1
  api_key: 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
  temperature: 0.0
  
  # Standard settings
  max_steps: 16
  temperature: 0.0

evaluation:
  enabled: true
  gold_root: kdd-benchmark/output
  run_after_execution: true
  fail_if_ground_truth_missing: true
  lambda_penalty: 0.1
  mode: standard

run:
  output_dir: kdd-benchmark/artifacts/runs
  run_id: 
  max_workers: 1
  task_timeout_seconds: 600

logging:
  log_debug: true  # Set to true to see debug logs