MethodError: no method matching ReinforcementLearningCore.TabularApproximator(::Vector{Float64}, ::Optimisers.Descent{Float64})

[D3MZ]

Error with Chapter 06 example
https://github.com/JuliaReinforcementLearning/ReinforcementLearningAnIntroduction.jl/blob/master/notebooks/Chapter06_RandomWalk.jl

MethodError: no method matching ReinforcementLearningCore.TabularApproximator(::Vector{Float64}, ::Optimisers.Descent{Float64})
The type `ReinforcementLearningCore.TabularApproximator` exists, but no method is defined for this combination of argument types when trying to construct it.

Closest candidates are:
  ReinforcementLearningCore.TabularApproximator(::A) where A<:AbstractArray
   @ ReinforcementLearningCore ~/.julia/packages/ReinforcementLearningCore/jIECU/src/policies/learners/tabular_approximator.jl:19
  ReinforcementLearningCore.TabularApproximator(::A) where A
   @ ReinforcementLearningCore ~/.julia/packages/ReinforcementLearningCore/jIECU/src/policies/learners/tabular_approximator.jl:4