Exploring the Limits of Weakly Supervised Pretraining

Open paper
Benchmark Model Rank Results
image-classification-on-imagenetResNeXt-101 32x48d#222Top 1 Accuracy: 85.4%Number of params: 829MGFLOPs: 306
image-classification-on-imagenetResNeXt-101 32x32d#247Top 1 Accuracy: 85.1%Number of params: 466MGFLOPs: 174
image-classification-on-imagenetResNeXt-101 32×16d#324Top 1 Accuracy: 84.2%Number of params: 194MGFLOPs: 72
image-classification-on-imagenetResNeXt-101 32x8d#541Top 1 Accuracy: 82.2%Number of params: 88M

Rank counts only results with a code link.