responses_create_params: temperature: 1.0 top_p: 0.95 max_output_tokens: 32768 # Mark runs unhealthy for excessive failures, including sidecar failures; keep partial scores. health_check_max_not_final_failure_fraction: 0.01 health_check_max_final_failure_fraction: 0.01 genrm_eval: resources_servers: genrm_eval: entrypoint: app.py domain: rlhf verified: false scoring_module: resources_servers.genrm_eval.preferences metric_key: overall/accuracy description: Evaluate a generative reward model against chat preferences genrm_eval_simple_agent: responses_api_agents: simple_agent: entrypoint: app.py max_steps: 1 resources_server: type: resources_servers name: genrm_eval model_server: type: responses_api_models name: policy_model datasets: - name: example type: example jsonl_fpath: resources_servers/genrm_eval/data/example.jsonl genrm_eval_environment_server: environment_servers: single_agent_turn_legacy: entrypoint: app.py resources_server: type: resources_servers name: genrm_eval agent_server: type: responses_api_agents name: genrm_eval_simple_agent resources_tool_transports: - direct_http