math_with_code: resources_servers: math_with_code: entrypoint: app.py max_execution_time: 10 domain: math verified: false description: Model solves competitive math problems using simple calculator tools value: Improve math and simple tool use capabilities math_with_code_simple_agent: responses_api_agents: simple_agent: entrypoint: app.py resources_server: type: resources_servers name: math_with_code model_server: type: responses_api_models name: policy_model datasets: - name: train type: train jsonl_fpath: resources_servers/math_with_code/data/train.jsonl gitlab_identifier: dataset_name: open_math_reasoning_problems_tool version: 0.0.1 artifact_fpath: open_math_reasoning_problems_tool.jsonl license: Apache 2.0 - name: example type: example jsonl_fpath: resources_servers/math_with_code/data/example.jsonl max_steps: 3 # Tighter cap to reduce long-tail rollout stragglers