CSPResNeXt-50

Wang et al., 2020 — CSPNet: A New Backbone that can Enhance Learning Capability of CNN (arXiv:1911.11929)

Lucid port of timm/cspresnext50.ra_in1k, converted to Lucid-native safetensors.

Available weights

Tag	acc@1	acc@5	Params	GFLOPs	Size	Source
RA_IN1K (default)	80.04	—	20.6M	—	78.72 MB	timm

Usage

import lucid.models as models
from lucid.models.weights import CSPResNeXt50Weights

# default tag
model = models.cspresnext_50_cls(pretrained=True)

# explicit tag (enum or string)
model = models.cspresnext_50_cls(weights=CSPResNeXt50Weights.RA_IN1K)
model = models.cspresnext_50_cls(pretrained="RA_IN1K")

# preprocessing travels with the weights
weights = CSPResNeXt50Weights.RA_IN1K
preprocess = weights.transforms()
logits = model(preprocess(image)[None]).logits

Conversion

Converted from timm/cspresnext50.ra_in1k via python -m tools.convert_weights cspresnext_50 --tag RA_IN1K. Key mapping + numerical parity verified against the source.

License

apache-2.0 — inherited from the original weights.

Citation

@inproceedings{wang2020cspnet,
  title={CSPNet: A New Backbone that can Enhance Learning Capability of CNN},
  author={Wang, Chien-Yao and Liao, Hong-Yuan Mark and Wu, Yueh-Hua and Chen, Ping-Yang and Hsieh, Jun-Wei and Yeh, I-Hau},
  booktitle={CVPR Workshops}, year={2020}
}