A lightweight multilabel classifier trained on an arbitrary subset of nyanko7/danbooru2023. Trained for ~40M samples at native resolution ~238x238. Decent performance for its size.
- Downloads last month
- 5
A lightweight multilabel classifier trained on an arbitrary subset of nyanko7/danbooru2023. Trained for ~40M samples at native resolution ~238x238. Decent performance for its size.
import timm model = timm.create_model("hf_hub:gustproof/dnbr-tagger-preview1", pretrained=True)