LightlyTrain_EX

How to Run

1. set up a virtual environment.

   conda create -n ltrain -y python=3.11
   conda activate ltrain
   
   pip install torch torchvision --index-url https://download.pytorch.org/whl/cu129
   pip install wandb
   pip install "lightly-train[wandb,timm]"
   pip install datasets matplotlib opencv-python
   

2. prepare imagenet100 dataset

   mkdir dataset
   wget https://huggingface.co/datasets/ilee0022/ImageNet100/resolve/main/label2text.json -P dataset
   

3. train resnet18 with imagenet100 dataset

   python train.py
   // 'best_model.pth' will be generated in checkpoint directory.
   

4. check a trained resnet18 with imagenet100 dataset

   python infer.py