FiFdZddlZddlmZGddejZy)zActivation modules.Nc^eZdZdZddfd Zdej dej fdZxZS)AGLUaFUnified activation function module from AGLU. This class implements a parameterized activation function with learnable parameters lambda and kappa, based on the AGLU (Adaptive Gated Linear Unit) approach. Attributes: act (nn.Softplus): Softplus activation function with negative beta. lambd (nn.Parameter): Learnable lambda parameter initialized with uniform distribution. kappa (nn.Parameter): Learnable kappa parameter initialized with uniform distribution. Methods: forward: Compute the forward pass of the Unified activation function. Examples: >>> import torch >>> m = AGLU() >>> input = torch.randn(2) >>> output = m(input) >>> print(output.shape) torch.Size([2]) References: https://github.com/kostas1515/AGLU returnc t|tjd|_tj tj jtjd|||_ tj tj jtjd|||_ y)zEInitialize the Unified activation function with learnable parameters.g)beta)devicedtypeN) super__init__nnSoftplusact Parameterinituniform_torchemptylambdkappa)selfr r __class__s c/var/www/pod-logistic/pod-ai/venv/lib/python3.12/site-packages/ultralytics/nn/modules/activation.pyr z AGLU.__init__"sx ;;D)\\"''"2"25;;qW\3]"^_ \\"''"2"25;;qW\3]"^_ xc tj|jd}tjd|z |j |j |ztj |z zS)a Apply the Adaptive Gated Linear Unit (AGLU) activation function. This forward method implements the AGLU activation function with learnable parameters lambda and kappa. The function applies a transformation that adaptively combines linear and non-linear components. Args: x (torch.Tensor): Input tensor to apply the activation function to. Returns: (torch.Tensor): Output tensor after applying the AGLU activation function, with the same shape as the input. g-C6?)minr)rclamprexprrlog)rrlams rforwardz AGLU.forward)sNkk$**&1yy!c'TXXtzzA~3.O%PPQQr)NN)rN) __name__ __module__ __qualname____doc__r rTensorr" __classcell__)rs@rrrs,2` R R%,, Rrr)r&rtorch.nnr Modulerrrr,s  .R299.Rr