LeNet¶

ConvNet Image Classification

class lucid.models.LeNet(conv_layers: list[dict], clf_layers: list[int], clf_in_features: int, _base_activation: ~typing.Type[~lucid.nn.module.Module] = <class 'lucid.nn.modules.activation.Tanh'>)¶

Overview¶

The LeNet base class provides a flexible implementation for defining various versions of the LeNet architecture, including LeNet-1, LeNet-4, and LeNet-5.

It allows the configuration of convolutional and fully connected layers through arguments, making it adaptable for different use cases.

Class Signature¶

class LeNet(nn.Module):
    def __init__(
        self,
        conv_layers: list[dict],
        clf_layers: list[int],
        clf_in_features: int,
     ) -> None

Parameters¶

conv_layers (list[dict]) A list of dictionaries specifying the configuration of the convolutional layers. Each dictionary should define the number of output channels (out_channels) and optionally other parameters such as kernel size, stride, and padding.
clf_layers (list[int]) A list specifying the sizes of fully connected (classifier) layers. Each entry represents the number of units in the respective layer.
clf_in_features (int) The number of input features for the first fully connected layer. This is determined by the output size of the feature extractor.

Attributes¶

feature_extractor (nn.Sequential) A sequential model containing the convolutional and pooling layers.
classifier (nn.Sequential) A sequential model containing the fully connected layers.

Methods¶

forward(x: Tensor) -> Tensor Performs the forward pass through the feature extractor and classifier.

def forward(self, x):
    x = self.feature_extractor(x)
    x = x.view(x.shape[0], -1)  # Flatten
    x = self.classifier(x)
    return x

Example Usage¶

Below is an example of defining and using a LeNet-based architecture:

import lucid.models as models

# Define a custom LeNet architecture
custom_lenet = models.LeNet(
    conv_layers=[
        {"out_channels": 6},
        {"out_channels": 16},
    ],
    clf_layers=[120, 84, 10],
    clf_in_features=16 * 5 * 5,
)

# Sample input tensor (e.g., 32x32 grayscale image)
input_tensor = lucid.Tensor([...])

# Forward pass
output = custom_lenet(input_tensor)
print(output)