metadata
library_name: lucid
license: bsd-3-clause
tags:
- image-classification
- convnext
- lucid
datasets:
- imagenet-1k
pipeline_tag: image-classification
model-index:
- name: convnext-small
results:
- task:
type: image-classification
dataset:
name: ImageNet-1K
type: imagenet-1k
metrics:
- type: acc@1
value: 83.616
- type: acc@5
value: 96.65
ConvNeXt-Small
Liu et al., 2022 — A ConvNet for the 2020s (arXiv:2201.03545)
Lucid port of torchvision/ConvNeXt_Small_Weights.IMAGENET1K_V1,
converted to Lucid-native safetensors.
Available weights
| Tag | acc@1 | acc@5 | Params | GFLOPs | Size | Source |
|---|---|---|---|---|---|---|
IMAGENET1K_V1 (default) |
83.616 | 96.65 | 50.2M | 8.684 | 191.62 MB | torchvision |
Usage
import lucid.models as models
from lucid.models.weights import ConvNeXtSmallWeights
# default tag
model = models.convnext_small_cls(pretrained=True)
# explicit tag (enum or string)
model = models.convnext_small_cls(weights=ConvNeXtSmallWeights.IMAGENET1K_V1)
model = models.convnext_small_cls(pretrained="IMAGENET1K_V1")
# preprocessing travels with the weights
weights = ConvNeXtSmallWeights.IMAGENET1K_V1
preprocess = weights.transforms()
logits = model(preprocess(image)[None]).logits
Conversion
Converted from torchvision/ConvNeXt_Small_Weights.IMAGENET1K_V1 via
python -m tools.convert_weights convnext_small --tag IMAGENET1K_V1.
Key mapping + numerical parity verified against the source.
License
bsd-3-clause — inherited from the original weights.
Citation
@inproceedings{liu2022convnet,
title={A ConvNet for the 2020s},
author={Liu, Zhuang and Mao, Hanzi and Wu, Chao-Yuan and Feichtenhofer, Christoph and Darrell, Trevor and Xie, Saining},
booktitle={CVPR}, year={2022}
}