---
library_name: lucid
license: apache-2.0
tags:
  - image-classification
  - inception
  - lucid
datasets:
  - imagenet-1k
pipeline_tag: image-classification
model-index:
  - name: inception-next-tiny
    results:
      - task: { type: image-classification }
        dataset: { name: ImageNet-1k, type: imagenet-1k }
        metrics:
          - { type: acc@1, value: 82.3 }
          - { type: acc@5, value: 96.1 }
---

# InceptionNeXt-Tiny

> Yu et al., 2024 — *InceptionNeXt: When Inception Meets ConvNeXt* (arXiv:2303.16900)

[Lucid](https://github.com/ChanLumerico/lucid) port of `timm/inception_next_tiny.sail_in1k`,
converted to Lucid-native safetensors.

## Available weights

| Tag | acc@1 | acc@5 | Params | GFLOPs | Size | Source |
|---|---|---|---|---|---|---|
| `SAIL_IN1K` *(default)* | 82.3 | 96.1 | 28.1M | — | 107.11 MB | timm |

## Usage

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

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

# explicit tag (enum or string)
model = models.inception_next_tiny_cls(weights=InceptionNeXtTinyWeights.SAIL_IN1K)
model = models.inception_next_tiny_cls(pretrained="SAIL_IN1K")

# preprocessing travels with the weights
weights = InceptionNeXtTinyWeights.SAIL_IN1K
preprocess = weights.transforms()
logits = model(preprocess(image)[None]).logits
```

## Conversion

Converted from `timm/inception_next_tiny.sail_in1k` via
`python -m tools.convert_weights inception_next_tiny --tag SAIL_IN1K`.
Key mapping + numerical parity verified against the source.

## License

`apache-2.0` — inherited from the original weights.

## Citation

```
@inproceedings{yu2024inceptionnext,
  title={InceptionNeXt: When Inception Meets ConvNeXt},
  author={Yu, Weihao and Zhou, Pan and Yan, Shuicheng and Wang, Xinchao},
  booktitle={CVPR}, year={2024}
}
```