ViT-MoE-15B (Every-2)

Reported on 6 benchmarks across 2 tasks · 1 paper · 6 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision6 results

Image ClassificationonImageNet - 5-shot
Top 1 Accuracy· 2021-06-10
82.78
SOTA
Scaling Vision with Sparse Mixture of Experts arXiv:2106.05974
Image ClassificationonImageNet - 10-shot
Top 1 Accuracy· 2021-06-10
84.29
best: 84.6 (MAWS (ViT-6.5B))
SOTA
Scaling Vision with Sparse Mixture of Experts arXiv:2106.05974
Image ClassificationonImageNet - 1-shot
Top 1 Accuracy· 2021-06-10
68.66
SOTA
Scaling Vision with Sparse Mixture of Experts arXiv:2106.05974
Few-Shot Image ClassificationonImageNet - 5-shot
Top 1 Accuracy· 2021-06-10
82.78
SOTA
Scaling Vision with Sparse Mixture of Experts arXiv:2106.05974
Few-Shot Image ClassificationonImageNet - 10-shot
Top 1 Accuracy· 2021-06-10
84.29
best: 84.6 (MAWS (ViT-6.5B))
SOTA
Scaling Vision with Sparse Mixture of Experts arXiv:2106.05974
Few-Shot Image ClassificationonImageNet - 1-shot
Top 1 Accuracy· 2021-06-10
68.66
SOTA
Scaling Vision with Sparse Mixture of Experts arXiv:2106.05974