PyTorch

{
  "run": {
    "kind": "BenchmarkRun",
    "spec": {
      "status": "passed",
      "timing": {
        "latencyNs": {
          "median": 75152893
        },
        "primaryStatistic": "median"
      },
      "observedAt": "2024-10-09T12:29:20.000Z",
      "measurements": [
        {
          "unit": "ns",
          "value": 75152893,
          "metric": "latency",
          "statistic": "p20"
        },
        {
          "unit": "ns",
          "value": 75152893,
          "metric": "latency",
          "statistic": "p80"
        }
      ],
      "sourceNative": {
        "source": "liger-kernel-bench",
        "metrics": {
          "ms_20": 75.15289306640625,
          "ms_50": 75.15289306640625,
          "ms_80": 75.15289306640625
        },
        "benchmark": "fused_linear_jsd/forward",
        "externalId": "fused_linear_jsd/forward/torch/nvidia-h100-80gb-hbm3/hidden4096-tokens8192-vocab128256-bf16/2024-10-09-12-29-20"
      },
      "protocolDigest": "sha256:a79eaf4df66c97c6adc1fb2d09df78682155d1ea86652ae0b70199875bde7747",
      "workloadDigest": "sha256:9b48d963b90f809b6a9ff44695b4f96afc4a3abe22158bfe5eb160a6b7746e34",
      "environmentDigest": "sha256:2750ae381d6de582f1eb6b9008e0987463d19ede8af127d97d76c5ae09121393",
      "implementationDigest": "sha256:e2560a134df774ba570c6e62110a7de637b91c5ababdacf09bebb0e7b49ca688"
    },
    "metadata": {
      "name": "liger-fused-linear-jsd-forward-torch-nvidia-h100-80gb-hbm3-hidden4096-tokens8192-vocab128256-bf16-20",
      "title": "Fused linear + JSD · torch · forward",
      "labels": {
        "liger_version": "0.3.1"
      }
    },
    "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 H100",
        "formFactor": "80GB HBM3",
        "architecture": "sm_90"
      },
      "software": {}
    },
    "metadata": {
      "name": "liger-bench-nvidia-h100-80gb-hbm3",
      "title": "Liger benchmark host · H100 80GB HBM3"
    },
    "apiVersion": "kernelindex.dev/v1alpha1"
  }
}