Transformers
Rotary position embedding · seq = 4096 · hidden = 8192 · q_heads = 32 · kv_heads = 8 · bf16 · run 01a016ae-52be…
●passedReported evidence
Not re-observed recently.
Primary measurement
2.18msmedian
Rank 2 in its comparison group · source-native comparison · observed 2024-09-03
ms 20 2.175136089324951 · ms 50 2.176192045211792 · ms 80 2.177433729171753
Identity
implementationTransformers
projectHugging Face Transformers
revisionunknown
workloadseq = 4096 · hidden = 8192 · q_heads = 32 · kv_heads = 8 · bf16
comparison keysha256:59c8aad0307ecba7…
sourceLiger-Kernel benchmarks
external idrope/full/huggingface/nvidia-a100-sxm4-80gb/hidden8192-kv_heads8-q_heads32-seq4096-bf16/2024-09-03-15-40-14
sha256:5b40d054c76fc0bd5898d9a…
Correctness
Marked passed by the source; the correctness policy was not published.
Workload
seq4096
hidden8192
q_heads32
kv_heads8
kbf16 [1, 8, 4096, 256]
qbf16 [1, 32, 4096, 256]
definition comparatornot_asserted
Measurements
latency · median2.18 ms
latency · p202.18 ms
latency · p802.18 ms
Protocol
harnessLiger-Kernel benchmark scripts
timercuda_events
primaryStatisticmedian
comparabilityFamilyliger_kernel_bench
Environment
gpuNVIDIA A100 (sm_80)
Artifacts
No artifacts published with this run.
Replications and notes
No attestations yet.
Community attestations. They never change the evidence level; only a KernelIndex-controlled rerun does.
Add a reproduction or note
Canonical manifest
Show manifestHide manifest
{
"run": {
"kind": "BenchmarkRun",
"spec": {
"status": "passed",
"timing": {
"latencyNs": {
"median": 2176192
},
"primaryStatistic": "median"
},
"observedAt": "2024-09-03T15:40:14.000Z",
"measurements": [
{
"unit": "ns",
"value": 2175136,
"metric": "latency",
"statistic": "p20"
},
{
"unit": "ns",
"value": 2177434,
"metric": "latency",
"statistic": "p80"
}
],
"sourceNative": {
"source": "liger-kernel-bench",
"metrics": {
"ms_20": 2.175136089324951,
"ms_50": 2.176192045211792,
"ms_80": 2.177433729171753
},
"benchmark": "rope/full",
"externalId": "rope/full/huggingface/nvidia-a100-sxm4-80gb/hidden8192-kv_heads8-q_heads32-seq4096-bf16/2024-09-03-15-40-14"
},
"protocolDigest": "sha256:5b02ee9a2140fbc6fcc1c89800f6c177ba22718c8897cb532bf0ba91867c1e58",
"workloadDigest": "sha256:29ee69f1665e3543c77a8d37b49448e7e1ed1f1d9a8a1cde0178b8e117a1e503",
"environmentDigest": "sha256:90c9e2764222f592477cfccf3373519add5a1b881563956ad058c4f99e41b6b9",
"implementationDigest": "sha256:459baf6556e5536285c26b26d11eda4980b44d362142461d8e7e5b4f3900ce23"
},
"metadata": {
"name": "liger-rope-full-huggingface-nvidia-a100-sxm4-80gb-hidden8192-kv-heads8-q-heads32-seq4096-bf16-2024-0",
"title": "Rotary position embedding · huggingface · full",
"labels": {
"liger_version": "0.2.1"
}
},
"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 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"
}
}Cite this record (permalink, digest, access date)
Report an issue with this run
Published 2026-08-18 · Liger-Kernel benchmarks · BSD-2-ClauseAll results for Rotary position embedding →JSON