convnext_tiny_cls(pretrained: bool | str = False, weights: ConvNeXtTinyWeights | None = None, overrides: object = {})ConvNeXt-Tiny image classifier (Liu et al., 2022).
Combines the convnext_tiny backbone with a global average
pool + LayerNorm + single nn.Linear classification head.
Default output is num_classes=1000 (ImageNet-1k). ~29M
parameters.
Model Size
Parameters
pretrainedbool= FalseIf
True, loads ImageNet-1k pretrained weights when
available. Defaults to False.Explicit weights enum member; takes precedence over
pretrained.**overridesobject= {}Keyword overrides on top of the canonical ConvNeXt-T config.
Returns
ConvNeXtForImageClassificationClassifier returning ImageClassificationOutput whose
logits has shape (B, num_classes).
Notes
ConvNeXt-T reaches 82.1% top-1 on ImageNet-1k (Liu et al., 2022, Table 1).
Examples
>>> import lucid
>>> from lucid.models.vision.convnext import convnext_tiny_cls
>>> model = convnext_tiny_cls(num_classes=1000)
>>> x = lucid.randn(1, 3, 224, 224)
>>> model(x).logits.shape
(1, 1000)
Load ImageNet-pretrained weights:
>>> model = convnext_tiny_cls(pretrained=True) # DEFAULT tag
>>> from lucid.models.weights import ConvNeXtTinyWeights
>>> model = convnext_tiny_cls(weights=ConvNeXtTinyWeights.IMAGENET1K_V1)