change kappa back to lambda

src/rules.jl

@@ -575,30 +575,30 @@ function adjust(r::WeightDecay; gamma = nothing, kw...)
  end
 
 """
-    SignDecay(κ = 1e-3)
+    SignDecay(λ = 1e-3)
 
 Implements ``L_1`` regularisation, also known as LASSO regression,
 when composed  with other rules as the first transformation in an [`OptimiserChain`](@ref).
 
-It does this by adding `κ .* sign(x)` to the gradient. This is equivalent to adding 
-`κ * sum(abs, x) == κ * norm(x, 1)` to the loss.
+It does this by adding `λ .* sign(x)` to the gradient. This is equivalent to adding 
+`λ * sum(abs, x) == λ * norm(x, 1)` to the loss.
 
 See also [`WeightDecay`] for ``L_2`` normalisation.
 They can be used together: `OptimiserChain(SignDecay(0.012), WeightDecay(0.034), Adam())`
 is equivalent to adding `0.012 * norm(x, 1) + 0.017 * norm(x, 2)^2` to the loss function.
 
 # Parameters
-- Penalty (`κ ≥ 0`): Controls the strength of the regularisation.
+- Penalty (`λ ≥ 0`): Controls the strength of the regularisation.
 """
 @def struct SignDecay <: AbstractRule
-  kappa = 1e-3
+  lambda = 1e-3
 end
 
 init(o::SignDecay, x::AbstractArray) = nothing
 
 function apply!(o::SignDecay, state, x::AbstractArray{T}, dx) where T
-  κ = T(o.kappa)
-  dx′ = @lazy dx + κ * sign(x)
+  λ = T(o.lambda)
+  dx′ = @lazy dx + λ * sign(x)
 
   return state, dx′
 end