Liger

{
  "run": {
    "kind": "BenchmarkRun",
    "spec": {
      "status": "passed",
      "timing": {
        "latencyNs": {
          "median": 14654464
        },
        "primaryStatistic": "median"
      },
      "observedAt": "2025-06-27T09:21:16.000Z",
      "measurements": [
        {
          "unit": "ns",
          "value": 14633984,
          "metric": "latency",
          "statistic": "p20"
        },
        {
          "unit": "ns",
          "value": 14680064,
          "metric": "latency",
          "statistic": "p80"
        }
      ],
      "sourceNative": {
        "source": "liger-kernel-bench",
        "metrics": {
          "ms_20": 14.63398380279541,
          "ms_50": 14.654463768005371,
          "ms_80": 14.68006420135498
        },
        "benchmark": "distill_cosine_loss/full",
        "externalId": "distill_cosine_loss/full/liger/nvidia-a100-sxm4-80gb/hidden4096-tokens1024-vocab128256-bf16/2025-06-27-09-21-16"
      },
      "protocolDigest": "sha256:5b02ee9a2140fbc6fcc1c89800f6c177ba22718c8897cb532bf0ba91867c1e58",
      "workloadDigest": "sha256:88ce4cf139b171c4dde8f7905153df3ff5ef235c7f81fee5d4dbfadfbe2569de",
      "environmentDigest": "sha256:90c9e2764222f592477cfccf3373519add5a1b881563956ad058c4f99e41b6b9",
      "implementationDigest": "sha256:0c28491dc6490ae8c096c5c517a08be0845b6109caadd451dd4036060ce43f3d"
    },
    "metadata": {
      "name": "liger-distill-cosine-loss-full-liger-nvidia-a100-sxm4-80gb-hidden4096-tokens1024-vocab128256-bf16-20",
      "title": "Chunked distillation cosine loss · liger · full",
      "labels": {
        "liger_version": "0.5.10"
      }
    },
    "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 forward pass plus the backward 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-full",
      "title": "Liger benchmark harness · full pass"
    },
    "apiVersion": "kernelindex.dev/v1alpha1"
  },
  "environment": {
    "kind": "ExecutionEnvironment",
    "spec": {
      "hardware": {
        "vendor": "nvidia",
        "product": "NVIDIA A100",
        "formFactor": "SXM4-80GB",
        "architecture": "sm_80"
      },
      "software": {}
    },
    "metadata": {
      "name": "liger-bench-nvidia-a100-sxm4-80gb",
      "title": "Liger benchmark host · A100-SXM4-80GB"
    },
    "apiVersion": "kernelindex.dev/v1alpha1"
  }
}