{
  "schema": "computearena-public-benchmark/1",
  "submission_id": "75aba215-6f95-49a7-a026-d876e7e88b4d",
  "run_id": "aa68acd9a59f41e8178b4f6a99050ab5",
  "created_at_unix_ms": 1789838617575,
  "submitted_by": "prabod",
  "model": {
    "identity_resolution": "unresolved",
    "identity_verification": "unresolved",
    "name": "Llama-3.2-1B-Instruct-cuda-q8",
    "architecture": "llama",
    "quantization": "base_q4",
    "artifact": {
      "sha256": "cf3be554b160c49b1a4d896950ba1d07f6fbce2f5392071dd8a29b10884f2c24",
      "format": "base",
      "quantization_namespace": "basert",
      "quantization_scheme": "base_q4"
    }
  },
  "runtime": {
    "name": "basert",
    "version": "0.2.6"
  },
  "benchmark": {
    "mode": "text",
    "chip": "NVIDIA GB10",
    "backend": "cuda",
    "protocol": {
      "report_schema": "computearena-benchmark/1",
      "benchmark_schema": "basert-benchmark-harness/1",
      "telemetry_schema": "basert-telemetry/4",
      "throughput_schema": "computearena-throughput/3",
      "comparable": true,
      "context_isolation": "headline_then_per_prefill",
      "execution_layout": "headline_then_prefill_processes",
      "model_load_in_timing": false,
      "decode_initial_context_tokens": 1,
      "conditioning_schema": "basert-conditioning/1",
      "conditioning_mode": "warmup_only",
      "cooldown_enabled": false
    },
    "parameters": {
      "context_tokens": 16384,
      "decode_context_tokens": 1
    },
    "throughput": {
      "prefill": [
        {
          "tokens": 128,
          "tokens_per_second": 14362.350349040475
        },
        {
          "tokens": 256,
          "tokens_per_second": 25104.52217785199
        },
        {
          "tokens": 512,
          "tokens_per_second": 34295.79613780694
        },
        {
          "tokens": 1024,
          "tokens_per_second": 38847.68333740768
        },
        {
          "tokens": 2048,
          "tokens_per_second": 38841.62635000332
        },
        {
          "tokens": 4096,
          "tokens_per_second": 33970.72587665182
        },
        {
          "tokens": 8192,
          "tokens_per_second": 28685.17633557909
        },
        {
          "tokens": 16384,
          "tokens_per_second": 21513.96836273078
        }
      ],
      "decode": {
        "generated_tokens": 128,
        "tokens_per_second": 173.65582206242644
      }
    }
  }
}