{
  "label": "GLM-5.3-Flash UD-IQ3_XXS (120.4 GB) reasoning LOW",
  "tests": {
    "01-blockfall": {
      "error": "DNF: exceeded the 60 minute time limit (partial output kept: 23014 chunks, 77180 thinking chars, 0 answer chars)",
      "partial_thinking_chars": 77180,
      "partial_answer_chars": 0,
      "partial_chunks": 23014
    },
    "03-eruption": {
      "error": "DNF: exceeded the 60 minute time limit (partial output kept: 23043 chunks, 62462 thinking chars, 12045 answer chars)",
      "partial_thinking_chars": 62462,
      "partial_answer_chars": 12045,
      "partial_chunks": 23043
    },
    "04-the-ledger": {
      "wall_seconds": 475.7,
      "prompt_tokens": 902,
      "completion_tokens": 3884,
      "tok_s_overall": 8.16,
      "thinking_chars": 1132,
      "answer_chars": 9163,
      "thinking_share": 0.11,
      "html_bytes": 9151,
      "gen_tok_s": 8.348646549263682,
      "prefill_tok_s": 90.43378646770287,
      "ttft_est_seconds": 9.87,
      "generate_seconds": 465.1,
      "thinking_seconds_est": 51.2,
      "timings_raw": {
        "cache_n": 9,
        "prompt_n": 893,
        "prompt_ms": 9874.628,
        "prompt_per_token_ms": 11.057814109742441,
        "prompt_per_second": 90.43378646770287,
        "predicted_n": 3884,
        "predicted_ms": 465105.329,
        "predicted_per_token_ms": 119.77989415400464,
        "predicted_per_second": 8.348646549263682
      }
    },
    "05-blind-artist": {
      "wall_seconds": 364.8,
      "prompt_tokens": 190,
      "completion_tokens": 3115,
      "tok_s_overall": 8.54,
      "thinking_chars": 315,
      "answer_chars": 7567,
      "thinking_share": 0.04,
      "html_bytes": 7555,
      "gen_tok_s": 8.619149788409423,
      "prefill_tok_s": 56.991700937781125,
      "ttft_est_seconds": 3.18,
      "generate_seconds": 361.3,
      "thinking_seconds_est": 14.5,
      "timings_raw": {
        "cache_n": 9,
        "prompt_n": 181,
        "prompt_ms": 3175.901,
        "prompt_per_token_ms": 17.54641436464088,
        "prompt_per_second": 56.991700937781125,
        "predicted_n": 3115,
        "predicted_ms": 361288.535,
        "predicted_per_token_ms": 116.0207241490045,
        "predicted_per_second": 8.619149788409423
      }
    }
  },
  "vision": {},
  "size_gb_on_disk": 120.4,
  "load_seconds": 35.0,
  "warmup_seconds": 1.7,
  "mem_baseline_gb": 5.55,
  "mem_peak_gb": 120.46,
  "mem_model_gb_est": 114.91,
  "backend": "llama-server direct: ~/repo/tools/llama.cpp-glm5next/build/bin/llama-server",
  "server_timings": [
    {
      "task": 0,
      "prompt_ms": 857.28,
      "prompt_tokens": 15,
      "prefill_tok_s": 17.5,
      "eval_ms": 784.52,
      "eval_tokens": 8,
      "gen_tok_s": 8.92
    },
    {
      "task": 11
    },
    {
      "task": 23101
    },
    {
      "task": 46183,
      "prompt_ms": 9874.63,
      "prompt_tokens": 893,
      "prefill_tok_s": 90.43,
      "eval_ms": 465105.33,
      "eval_tokens": 3884,
      "gen_tok_s": 8.35
    },
    {
      "task": 50072,
      "prompt_ms": 3175.9,
      "prompt_tokens": 181,
      "prefill_tok_s": 56.99,
      "eval_ms": 361288.53,
      "eval_tokens": 3115,
      "gen_tok_s": 8.62
    }
  ]
}