Liger

{
  "run": {
    "kind": "BenchmarkRun",
    "spec": {
      "status": "passed",
      "timing": {
        "latencyNs": {
          "median": 1731584
        },
        "primaryStatistic": "median"
      },
      "observedAt": "2025-04-30T17:22:12.000Z",
      "measurements": [
        {
          "unit": "ns",
          "value": 1730496,
          "metric": "latency",
          "statistic": "p20"
        },
        {
          "unit": "ns",
          "value": 1815552,
          "metric": "latency",
          "statistic": "p80"
        }
      ],
      "sourceNative": {
        "source": "liger-kernel-bench",
        "metrics": {
          "ms_20": 1.7304960489273071,
          "ms_50": 1.7315839529037476,
          "ms_80": 1.815551996231079
        },
        "benchmark": "sparse_multi_token_attention/forward",
        "externalId": "sparse_multi_token_attention/forward/liger/nvidia-geforce-rtx-3090/batch2-heads_in4-heads_out4-k3-seq512-fp32/2025-04-30-17-22-12"
      },
      "protocolDigest": "sha256:a79eaf4df66c97c6adc1fb2d09df78682155d1ea86652ae0b70199875bde7747",
      "workloadDigest": "sha256:2f8be9b6ee9a376f5e2fc8a8a63b7e44179a7ccbc51f1aecc45ed2184053973c",
      "environmentDigest": "sha256:b12fdfab6a337672499b10212a1e986e49e4383777bd7cbb58321b09af4bf6a3",
      "implementationDigest": "sha256:e1ceb494ae4e51bbeb1012eaf8c69999502f3336c84232c2a9077090ae8afa05"
    },
    "metadata": {
      "name": "liger-sparse-multi-token-attention-forward-liger-nvidia-geforce-rtx-3090-batch2-heads-in4-heads-out4",
      "title": "Sparse multi-token attention · liger · forward",
      "labels": {
        "liger_version": "0.5.8"
      }
    },
    "apiVersion": "kernelindex.dev/v1alpha1"
  },
  "protocol": {
    "kind": "BenchmarkProtocol",
    "spec": {
      "harness": {
        "name": "Liger-Kernel benchmark scripts",
        "repository": "https://github.com/linkedin/Liger-Kernel"
      },
      "measurement": {
        "timer": "cuda_events",
        "primaryStatistic": "median"
      },
      "comparability": {
        "notes": "Timed: the module forward pass, via triton.testing.do_bench with quantiles 0.5/0.2/0.8 (median with p20/p80 spread). The upstream CSV records no CUDA, driver, or torch version; the Liger release rides each run's labels. Comparable only within one kernel, workload, timed pass, and GPU.",
        "family": "liger_kernel_bench"
      }
    },
    "metadata": {
      "name": "liger-bench-do-bench-forward",
      "title": "Liger benchmark harness · forward pass"
    },
    "apiVersion": "kernelindex.dev/v1alpha1"
  },
  "environment": {
    "kind": "ExecutionEnvironment",
    "spec": {
      "hardware": {
        "vendor": "nvidia",
        "product": "NVIDIA GeForce RTX 3090",
        "architecture": "sm_86"
      },
      "software": {}
    },
    "metadata": {
      "name": "liger-bench-nvidia-geforce-rtx-3090",
      "title": "Liger benchmark host · GeForce RTX 3090"
    },
    "apiVersion": "kernelindex.dev/v1alpha1"
  }
}