ProgResViT DeiT-S (160 → 384)
EMA inference checkpoint for the DeiT-S ProgResViT model trained on ImageNet-1K.
- Training: standard
- Paper-reported full-path top-1 accuracy: 83.700%
- Full-path compute: 16.152 GMACs
Download
from huggingface_hub import hf_hub_download
checkpoint = hf_hub_download(
repo_id="NCPS/progresvit-deit-s-160-384-imagenet1k",
filename="progresvit_160_384.pth.tar",
)
The file contains EMA weights and public architecture metadata only. Use it with the ProgResViT repository:
python validate.py /path/to/imagenet/val \
--config 160_384 \
--checkpoint checkpoints/progresvit_160_384.pth.tar