{
  "description": "Numeric summary of one ramped-arrival LFM3-1.7B FP8 run; no prompts, completions, or credentials are included. Rates are whole-run averages unless noted.",
  "run_created_at": "2026-10-28T17:15:52.952754+00:00",
  "gpu": "NVIDIA GeForce RTX 4060 Laptop GPU (8188 MiB)",
  "model": "lfm3-1.7b-fp8",
  "server_running_request_limit": 64,
  "requests": 128,
  "shared_system_prompt_words": 500,
  "arrival_spacing_s": 0.1,
  "thinking": true,
  "max_output_tokens_per_request": 2048,
  "wall_s_rounded": 64.252,
  "latency": {
    "ttft_s": {
      "p50": 4.861,
      "p95": 24.63,
      "p99": 25.012,
      "max": 25.043
    },
    "e2e_s": {
      "p50": 31.855,
      "p95": 52.368,
      "p99": 52.536,
      "max": 52.798
    },
    "chunk_interval_ms_p50_aggregate": 13.54,
    "max_in_flight": 128,
    "mean_in_flight": 75.55
  },
  "throughput": {
    "generated_tokens": 234344,
    "aggregate_generated_tok_s": 3647.2,
    "decode_window_avg_tok_s": 3660.8,
    "prompt_tokens_total": 89854,
    "cached_prompt_tokens_total": 63488,
    "cached_prompt_fraction": 0.7066
  },
  "server_counter_deltas": {
    "num_requests_total": 128.0,
    "generation_tokens_total": 234344.0,
    "prompt_tokens_total": 89854.0,
    "cached_tokens_total": 63488.0
  },
  "gpu_busy_numeric": {
    "busy_samples": 47,
    "graphics_mhz_median": 2700.0,
    "memory_mhz_median": 8201.0,
    "power_w_median": 97.87,
    "temperature_c_max": 83.0,
    "memory_used_mib_median": 7769.0
  },
  "scope_note": "Single run; not a matched steady-state comparison or quality evaluation. Reasoning tokens are included in generated tokens."
}