Some machine and env info:
Torch 1.5.0+cu101
Current device 0
Devices count: 1
Device: Tesla P100-PCIE-16GBThe provided table show comparative analysys of between ResNet and VGG models on Cifar10. Results are obtained for models, that trainded 300 epocs
| Name | # layers | # params | Prec@1 |
|---|---|---|---|
| ResNet20 | 20 | 0.27M | 91.99% |
| VGG11 | 11 | 133M | 86.81% |
| VGG16 (135 epc) | 16 | 138 | 91.86% |
Training process was monitored by using TensorBoard tools
- Update Table
- Link to disk
- Results for
ResNet32,ResNet44,ResNet56,ResNet110,ResNet120,VGG11_bn,VGG13,VGG13_bn,VGG16_bn,VGG19,VGG19_bn