Pre-training Vision Models with Mandelbulb Variations

Open paper
Benchmark Model Rank Results
image-classification-on-cifar-10ViT-B (MandelbulbVAR-Hybrid-21k)#69Percentage correct: 98.2
image-classification-on-cifar-100ViT-B (MandelbulbVAR-Hybrid-21k)#111Percentage correct: 83.6
image-classification-on-imagenetViT-B (MandelbulbVAR-Hybrid-21k)#606Top 1 Accuracy: 82.2%

Rank counts only results with a code link.