ResNet-50

He et al., 2015 — Deep Residual Learning for Image Recognition (arXiv:1512.03385)

Lucid port of torchvision/ResNet50_Weights.IMAGENET1K_V1, converted to Lucid-native safetensors.

Available weights

Tag	acc@1	acc@5	Params	GFLOPs	Size	Source
IMAGENET1K_V1 (default)	76.13	92.862	25.6M	4.089	97.72 MB	torchvision

Usage

import lucid.models as models
from lucid.models.weights import ResNet50Weights

# default tag
model = models.resnet_50_cls(pretrained=True)

# explicit tag (enum or string)
model = models.resnet_50_cls(weights=ResNet50Weights.IMAGENET1K_V1)
model = models.resnet_50_cls(pretrained="IMAGENET1K_V1")

# preprocessing travels with the weights
weights = ResNet50Weights.IMAGENET1K_V1
preprocess = weights.transforms()
logits = model(preprocess(image)[None]).logits

Conversion

Converted from torchvision/ResNet50_Weights.IMAGENET1K_V1 via python -m tools.convert_weights resnet_50 --tag IMAGENET1K_V1. Key mapping + numerical parity verified against the source.

License

bsd-3-clause — inherited from the original weights.

Citation

@inproceedings{he2016deep,
  title={Deep Residual Learning for Image Recognition},
  author={He, Kaiming and Zhang, Xiangyu and Ren, Shaoqing and Sun, Jian},
  booktitle={CVPR}, year={2016}
}