Ordinal logit/probit model with long time to run

I want to modelling an ordinal logit/probit model. I have 50000 samples and 20 variables, y=1,2,…,21
so, I want to use a simple model in rstan


data {
int<lower=2> K;
int<lower=0> N;
int<lower=1> D;
array[N] int<lower=1, upper=K> y;
matrix[N,D] x;
}
parameters {
vector[D] beta;
ordered[K - 1] c;
}
model {
y ~ ordered_logistic(x * beta, c);
}


But it’s too slow. And I got

Rejecting initial value:
Log probability evaluates to log(0), i.e. negative infinity.
Stan can't start sampling from this initial value.

in the sampling. really make me confuse
Some suggestions?

these type of models benefit a lot from brute-force within-chain parallelisation, which you get with reduce_sum (have a look at the docs for this). This is only available in newer Stan versions, which you can get with cmdstanr.