README.md

---
library_name: lucid
license: bsd-3-clause
tags:
  - image-classification
  - convnext
  - lucid
datasets:
  - imagenet-1k
pipeline_tag: image-classification
model-index:
  - name: convnext-base
    results:
      - task: { type: image-classification }
        dataset: { name: ImageNet-1K, type: imagenet-1k }
        metrics:
          - { type: acc@1, value: 84.062 }
          - { type: acc@5, value: 96.87 }
---

# ConvNeXt-Base

> Liu et al., 2022 — *A ConvNet for the 2020s* (arXiv:2201.03545)

[Lucid](https://github.com/ChanLumerico/lucid) port of `torchvision/ConvNeXt_Base_Weights.IMAGENET1K_V1`,
converted to Lucid-native safetensors.

## Available weights

| Tag | acc@1 | acc@5 | Params | GFLOPs | Size | Source |
|---|---|---|---|---|---|---|
| `IMAGENET1K_V1` *(default)* | 84.062 | 96.87 | 88.6M | 15.355 | 337.98 MB | torchvision |

## Usage

```python
import lucid.models as models
from lucid.models.weights import ConvNeXtBaseWeights

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

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

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

## Conversion

Converted from `torchvision/ConvNeXt_Base_Weights.IMAGENET1K_V1` via
`python -m tools.convert_weights convnext_base --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}
}
```