deepsearchqa: resources_servers: deepsearchqa: entrypoint: app.py domain: knowledge verified: false description: Grade DeepSearchQA single-answer and set-answer responses value: Evaluate multi-step web research answers judge_model_server: type: responses_api_models name: policy_model judge_responses_create_params: input: [] max_output_tokens: 32768 temperature: 0.0 top_p: 1.0 deepsearchqa_example_agent: responses_api_agents: harness_exa_search: entrypoint: app.py resources_server: type: resources_servers name: deepsearchqa model_server: type: responses_api_models name: policy_model datasets: - name: example type: example jsonl_fpath: resources_servers/deepsearchqa/data/example.jsonl deepsearchqa_example_environment_server: environment_servers: legacy_agent: entrypoint: app.py agent_server: type: responses_api_agents name: deepsearchqa_example_agent