speed_bench_resources_server: resources_servers: speed_bench: allowed_agents: [speed_bench_agent] entrypoint: app.py domain: other verified: false description: Speculative-decoding throughput benchmark. Aggregates vLLM per-request acceptance metrics returned by the agent. value: Measure inference-time speculative-decoding effectiveness for serving research and regression tests. metrics_source: response speed_bench_simple_agent: responses_api_agents: speed_bench_agent: entrypoint: app.py resources_server: type: resources_servers name: speed_bench_resources_server model_server: type: responses_api_models name: policy_model datasets: - name: example type: example jsonl_fpath: resources_servers/speed_bench/data/example.jsonl num_repeats: 1 license: NVIDIA Evaluation Dataset License Agreement speed_bench_simple_environment_server: environment_servers: legacy_agent: entrypoint: app.py agent_server: type: responses_api_agents name: speed_bench_simple_agent