lambda_1
This keyword sets the weight \(\lambda_1\) of the \(\mathcal{L}_1\)-norm regularization term in the loss function. The syntax is:
lambda_1 <weight>
Here, <weight>
represents \(\lambda_1\), which can be set to any non-negative values.
The default value is \(\lambda_1 = \sqrt{N}/1000\), where \(N\) is the total number of training parameters.