yjh0410 2a3974d075 add dl tutorial		vor 1 Jahr
..
data	2a3974d075 add dl tutorial	vor 1 Jahr
models	2a3974d075 add dl tutorial	vor 1 Jahr
utils	2a3974d075 add dl tutorial	vor 1 Jahr
.gitignore	2a3974d075 add dl tutorial	vor 1 Jahr
README.md	2a3974d075 add dl tutorial	vor 1 Jahr
engine.py	2a3974d075 add dl tutorial	vor 1 Jahr
main.py	2a3974d075 add dl tutorial	vor 1 Jahr
requirements.txt	2a3974d075 add dl tutorial	vor 1 Jahr

General Image Classification Laboratory

Train

For example, we are going to train ConvNet designed in this repo, so we can use the following command:

cd Vision-Pretraining-Tutorial/image_classification/
python main.py --cuda \
               --dataset cifar \
               --model convnet \
               --batch_size 256 \
               --optimizer adamw \
               --base_lr 1e-3 \
               --min_lr 1e-6

Evaluate

Evaluate the top1 & top5 accuracy:

cd Vision-Pretraining-Tutorial/image_classification/
python main.py --cuda \
           --dataset cifar \
           --model convnet \
           --batch_size 256 \
           --eval \
           --resume path/to/checkpoint

README.md

General Image Classification Laboratory

Train

Evaluate