PyTorch
python · BSD-3-Clause
holds 4 records
Use it
Source available upstream · BSD-3-ClauseView source →
No install recipe recorded for this revision.
interfacepython
Compatibility
measured onNVIDIA A100
declared hardwaredeclared only
architectures—
dtypes
Benchmark evidence
24 measurements across 1 GPU, fastest first.
Operation / workload
Hardware
Latency
Rank
Observed
GRPO loss, token-level importance samplingbf16 · [2, 1025, 128256]
NVIDIA A100
22.0ms
#3 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [2, 1025, 128256]
NVIDIA A100
22.0ms
#4 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [2, 1025, 128256]
NVIDIA A100
23.7ms
#1 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [2, 1025, 128256]
NVIDIA A100
23.7ms
#2 of 4
2025-08-04
GRPO loss, token-level importance samplingbf16 · [4, 1025, 128256]
NVIDIA A100
41.5ms
#2 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [4, 1025, 128256]
NVIDIA A100
41.8ms
#3 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [2, 1025, 128256]
NVIDIA A100
45.4ms
#3 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [2, 1025, 128256]
NVIDIA A100
46.0ms
#4 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [4, 1025, 128256]
NVIDIA A100
46.9ms
#1 of 4
2025-08-04
GRPO loss, token-level importance samplingbf16 · [4, 1025, 128256]
NVIDIA A100
47.4ms
#2 of 4
2025-08-05
Show all 24 measurements ›Showing all 24 measurements ⌄
GRPO loss, token-level importance samplingbf16 · [8, 1025, 128256]
NVIDIA A100
81.2ms
#2 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [8, 1025, 128256]
NVIDIA A100
81.7ms
#3 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [4, 1025, 128256]
NVIDIA A100
88.6ms
#3 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [4, 1025, 128256]
NVIDIA A100
90.0ms
#4 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [8, 1025, 128256]
NVIDIA A100
94.5ms
#1 of 4
2025-08-04
GRPO loss, token-level importance samplingbf16 · [8, 1025, 128256]
NVIDIA A100
94.8ms
#2 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [16, 1025, 128256]
NVIDIA A100
160.8ms
#3 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [16, 1025, 128256]
NVIDIA A100
162.6ms
#4 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [8, 1025, 128256]
NVIDIA A100
176.9ms
#3 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [8, 1025, 128256]
NVIDIA A100
177.2ms
#4 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [16, 1025, 128256]
NVIDIA A100
190.0ms
#1 of 4
2025-08-04
GRPO loss, token-level importance samplingbf16 · [16, 1025, 128256]
NVIDIA A100
190.1ms
#2 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [16, 1025, 128256]
NVIDIA A100
356.0ms
#3 of 4
2025-08-05
GRPO loss, token-level importance samplingbf16 · [16, 1025, 128256]
NVIDIA A100
356.5ms
#4 of 4
2025-08-05
Reported · How evidence levels are derived →
Source and license
sourceavailable
revision digestsha256:39aefd5bbcd914f3146a525b2dd28e8598242e7015bd3eeb1a37e7217fac6484
license declaredBSD-3-Clause
license concludedBSD-3-Clause
imported2026-08-20
Best evidence level for this revision: reported
JSON