List your company
ResearchFormerly WizWand

Inference Latency on ImageNet-1K 1.0 (val)

4.4Latency (ms)EffNet-B3
3.48289.673915.86522.0561Jul 3, 2026
Updated 3mo ago

Evaluation results

MethodLinks
EffNet-B3Model Scale=Micro, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.074.4
VNCT-MicroModel Scale=Micro, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.075.25
VSSD-MModel Scale=Micro, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.075.8
ConvNeXt-TModel Scale=Tiny, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.077.27
VNCT-TinyModel Scale=Tiny, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.077.31
VSSD-TModel Scale=Tiny, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.078.01
Swin-TModel Scale=Tiny, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.078.43
VMamba-TModel Scale=Tiny, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.0710.09
VNCT-SmallModel Scale=Small, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.0712.35
ConvNeXt-SModel Scale=Small, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.0713.03
VSSD-SModel Scale=Small, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.0713.24
VMamba-SModel Scale=Small, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.0718.31
ConvNeXt-BModel Scale=Base, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.0721.25
VNCT-BaseModel Scale=Base, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.0722.1
VSSD-BModel Scale=Base, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.0723.19
VMamba-BModel Scale=Base, Hardware=NVIDIA RTX 6000 Ada GPU, Precision=FP16, Batch size=1, Resolution=224 x 2242026.0727.33