mobilenet_v4_hybrid_medium¶

lucid.models.mobilenet_v4_hybrid_medium(num_classes: int = 1000, **kwargs) → MobileNet_V4¶

Overview¶

The mobilenet_v4_hybrid_medium function returns an instance of the MobileNet_V4 model featuring a hybrid architecture that blends traditional convolutional layers with innovative enhancements. This medium variant is optimized to balance computational efficiency with improved accuracy, making it suitable for a wide range of applications.

Total Parameters: 11,070,136

Function Signature¶

@register_model
def mobilenet_v4_hybrid_medium(num_classes: int = 1000, **kwargs) -> MobileNet_V4

Parameters¶

num_classes (int, optional): Defines the number of output classes. The default is 1000, commonly used for large-scale classification tasks.
**kwargs (dict): Additional keyword arguments that allow customization of the hybrid configuration. These parameters can be used to fine-tune the model’s architecture based on specific requirements.

Usage Example¶

>>> import lucid.models as models
>>> model = models.mobilenet_v4_hybrid_medium(num_classes=1000)
>>> print(model)

Details¶

The mobilenet_v4_hybrid_medium function constructs a MobileNet-v4 model with a hybrid architecture that combines the benefits of convolutional layers with enhanced module designs. It provides a compromise between accuracy and computational cost, making it a versatile choice for various deployment scenarios.

Note

This medium hybrid variant is ideal for users seeking a balance between performance and resource consumption. For specific needs, consider adjusting the configuration parameters via the additional keyword arguments.