{
  "schema": "computearena-public-benchmark/1",
  "submission_id": "65331601-0e96-47d0-8780-6166a39c3675",
  "run_id": "cb60e570667495cf1527eddc091903e7",
  "created_at_unix_ms": 1791555143886,
  "submitted_by": "basecompute",
  "model": {
    "canonical_id": "Qwen/Qwen3-4B",
    "identity_resolution": "artifact_verified",
    "identity_verification": "verified",
    "verified_at": "2026-09-11T12:18:35.929Z",
    "name": "Qwen3-4B-Q4",
    "architecture": "qwen",
    "quantization": "base_q4",
    "artifact": {
      "provider": "huggingface",
      "repo_id": "basecompute/Qwen3-4B",
      "revision": "7a98a7a34a8e0d1adf4a610fd8605b48d8beff77",
      "path": "Qwen3-4B-Q4.base",
      "sha256": "df660a56b61bb68b940da100f9ad345a93d396df1955387e2f41f317e130e696",
      "format": "base",
      "quantization_namespace": "basert",
      "quantization_scheme": "base_q4"
    }
  },
  "runtime": {
    "name": "basert",
    "version": "0.3.0"
  },
  "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": 2142.3662285751616
        },
        {
          "tokens": 256,
          "tokens_per_second": 2581.922900687234
        },
        {
          "tokens": 512,
          "tokens_per_second": 2841.5104792109128
        },
        {
          "tokens": 1024,
          "tokens_per_second": 2856.7705867941454
        },
        {
          "tokens": 2048,
          "tokens_per_second": 2878.9521938578805
        },
        {
          "tokens": 4096,
          "tokens_per_second": 2735.9703029127922
        },
        {
          "tokens": 8192,
          "tokens_per_second": 2611.24972472153
        },
        {
          "tokens": 16384,
          "tokens_per_second": 2340.1402940013872
        }
      ],
      "decode": {
        "generated_tokens": 128,
        "tokens_per_second": 89.41934038402388
      }
    }
  }
}