Specifying loc and scale on beta prior requires bounds, gives error

Sorry, I thought that you had an original model that threw the initial point failure error and then tried to bound the beta distribution and got “multiple values for keyword argument ‘transform” error. Is that code the first or the second?