Seer

DINOv3 ViT-L AI-generated image detector (~302M params) for TikTok TechJam 2026.

This repo hosts best.pt from the seer_vitl training run.

Load

from huggingface_hub import hf_hub_download

path = hf_hub_download("glennwuwu/seer", "best.pt")

Then load with the Seer trainer/infer code:

uv run python main.py infer --checkpoint best.pt --image photo.jpg

Downloads last month: -; Downloads are not tracked for this model. How to track