Skip to content
KernelIndex
Search⌘K

Transformers

Hugging Face Transformers · python · Apache-2.0
holds 2 records

Use it

Source available upstream · Apache-2.0View source →

No install recipe recorded for this revision.

interfacepython
Compatibility
measured onNVIDIA GeForce RTX 4090, NVIDIA A100
declared hardwaredeclared only
architectures—
dtypes

Benchmark evidence

34 measurements across 2 GPUs, fastest first.

Operation / workload
Hardware
Latency
Rank
Observed
SwiGLU MLPbf16 · [2, 1024, 2048]
NVIDIA GeForce RTX 4090
674.8µs
#2 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 1024, 2048]
NVIDIA GeForce RTX 4090
745.5µs
#4 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 2048, 2048]
NVIDIA GeForce RTX 4090
1.41ms
#3 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 2048, 2048]
NVIDIA GeForce RTX 4090
1.43ms
#4 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 1024, 2048]
NVIDIA GeForce RTX 4090
1.55ms
#2 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 1024, 2048]
NVIDIA GeForce RTX 4090
2.25ms
#2 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 1024, 2048]
NVIDIA GeForce RTX 4090
2.67ms
#4 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 4096, 2048]
NVIDIA GeForce RTX 4090
2.83ms
#3 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 4096, 2048]
NVIDIA GeForce RTX 4090
2.91ms
#4 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 2048, 2048]
NVIDIA GeForce RTX 4090
3.19ms
#2 of 4
2025-11-11
Show all 34 measurements ›
SwiGLU MLPbf16 · [2, 1024, 2048]
NVIDIA GeForce RTX 4090
3.43ms
#4 of 4
2025-11-11
SwiGLU MLPbf16 · [4, 1024, 4096]
NVIDIA A100
4.58ms
#1 of 2
2024-09-03
SwiGLU MLPbf16 · [2, 2048, 2048]
NVIDIA GeForce RTX 4090
4.59ms
#2 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 2048, 2048]
NVIDIA GeForce RTX 4090
4.74ms
#4 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 8192, 2048]
NVIDIA GeForce RTX 4090
5.66ms
#2 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 8192, 2048]
NVIDIA GeForce RTX 4090
6.08ms
#4 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 2048, 2048]
NVIDIA GeForce RTX 4090
6.16ms
#4 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 4096, 2048]
NVIDIA GeForce RTX 4090
6.27ms
#2 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 4096, 2048]
NVIDIA GeForce RTX 4090
9.08ms
#4 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 4096, 2048]
NVIDIA GeForce RTX 4090
9.23ms
#2 of 4
2025-11-11
SwiGLU MLPbf16 · [4, 2048, 4096]
NVIDIA A100
9.25ms
#1 of 2
2024-09-03
SwiGLU MLPbf16 · [2, 16384, 2048]
NVIDIA GeForce RTX 4090
11.3ms
#2 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 4096, 2048]
NVIDIA GeForce RTX 4090
11.9ms
#4 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 8192, 2048]
NVIDIA GeForce RTX 4090
12.1ms
#2 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 16384, 2048]
NVIDIA GeForce RTX 4090
12.4ms
#4 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 8192, 2048]
NVIDIA GeForce RTX 4090
17.9ms
#2 of 4
2025-11-11
SwiGLU MLPbf16 · [4, 4096, 4096]
NVIDIA A100
18.2ms
#2 of 2
2024-09-03
SwiGLU MLPbf16 · [2, 8192, 2048]
NVIDIA GeForce RTX 4090
18.7ms
#4 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 16384, 2048]
NVIDIA GeForce RTX 4090
23.9ms
#2 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 8192, 2048]
NVIDIA GeForce RTX 4090
24.8ms
#4 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 16384, 2048]
NVIDIA GeForce RTX 4090
35.3ms
#2 of 4
2025-11-11
SwiGLU MLPbf16 · [4, 8192, 4096]
NVIDIA A100
36.3ms
#2 of 2
2024-09-03
SwiGLU MLPbf16 · [2, 16384, 2048]
NVIDIA GeForce RTX 4090
37.1ms
#4 of 4
2025-11-11
SwiGLU MLPbf16 · [2, 16384, 2048]
NVIDIA GeForce RTX 4090
49.6ms
#4 of 4
2025-11-11

Reported · How evidence levels are derived →

Source and license

sourceavailable
revision digestsha256:70659a76da921ebc2777c752d1b601f56545c50a49056f176d2bf35d29374512
license declaredApache-2.0
license concludedApache-2.0
imported2026-08-18

Best evidence level for this revision: reported

JSON