FanCy-Learning

Fang Chen's Deep Learning Toolbox

Image Classification

Backbones & Datasets Zoo:

• Backbones
  • MobileNetV2 -- code link
  • MobileNetV3 (small, large) -- code link
  • ResNet & ResNeXt (with depth: 18, 34, 50, 101, 152) -- code link
  • Vision Transformer -- code link
  • SwinMLP & SwinTransformerV2 -- code link
• Datasets
  • Cifar10 / Cifar100
  • STL10
  • ImageNet

Running Command:

# train
> python ./image_classification/train_val.py --root ./ --seed 0 --devices 0 --dataset STL10 --im_size 128 --batch_size 8 --architecture mobilenetv2 --pretrained '' --epochs 300 --learning_rate 1e-2 --lr_interval '0.6 0.8 0.9' --lr_reduce 5 --optimizer SGD --log
# test only
> python ./image_classification/train_val.py --root ./ --seed 0 --devices 0 --dataset STL10 --im_size 128 --batch_size 8 --architecture mobilenetv2 --pretrained '' --epochs 300 --learning_rate 1e-2 --lr_interval '0.6 0.8 0.9' --lr_reduce 5 --optimizer SGD --log --test_only

Object Detection

coming soon