Dataset Overview

Datasets

• Office‑31:

  • Download link: Office‑31
  • Storage path: data/office/domain_adaptation_images/

• Office‑Home:

  • Download link: Office‑Home
  • Storage path: data/office-home/

• VisDA‑2017:

  • Download link: VisDA‑2017
  • Storage path: data/

• DomainNet:

  • Download link: DomainNet
  • Storage path: data/

Training

• Training Command:
  • Example:

    python3 main.py --train_batch_size 16 --dataset office --name wa \
        --source_list data/office/webcam_list.txt --target_list data/office/amazon_list.txt \
        --test_list data/office/amazon_list.txt --num_classes 31 --model_type ViT-B_16 \
        --pretrained_dir checkpoint/ViT-B_16.npz --num_steps 5000 --img_size 256 \
        --beta 0.1 --gamma 0.2 --use_im --theta 0.1
    

  • All commands are listed in script.txt.

Pre‑trained Models

• ViT‑B_16:
  • ImageNet‑21K: ViT‑B_16 (ImageNet‑21K)
  • ImageNet: ViT‑B_16 (ImageNet)
  • Storage path: checkpoint/