How to change the number of draws I want in my generated quantity

General
1

I have a super simple model. How do I generate more draws in my new generated quantities. The results are shown at the bottom of this post. It seems like the draws are set at 1000 by default. How can I increase or decrease this number?

stan.py

import cmdstanpy

import os
from cmdstanpy import CmdStanModel

import numpy as np

# naive model
naive_stan = os.path.join("naive_stan", "naive.stan")
out_dir = os.path.join("mean_stan", "out")
mean_model = CmdStanModel(stan_file=naive_stan)
d = np.random.normal(20, 1, size=100)
horizon = 3
cmdstanpy_data = {"horizon": horizon, "T": len(d), "y": d}
mean_fit = mean_model.sample(data=cmdstanpy_data, output_dir=out_dir, chains=1)
new_quantities = mean_model.generate_quantities(
    data=cmdstanpy_data, mcmc_sample=mean_fit
)
new_quantities.draws_pd()

naive.stan

data {
  int<lower=0> T;
  int<lower=0> horizon;
  vector[T] y;
}

parameters {
  real<lower=0> sigma;
}

model {
  y ~ normal(mean(y), sigma);
}

generated quantities {
  vector[horizon] forecast;
  forecast[1] = student_t_rng(T - 1, y[T], sigma * sqrt(1));
  for (h in 2:horizon){
    forecast[h] = student_t_rng(T - 1, forecast[h-1], sigma * sqrt(h));
  }

}

INFO:cmdstanpy:found newer exe file, not recompiling
INFO:cmdstanpy:compiled model file: /home/user/code/stan/tablespoon/naive_stan/naive
INFO:cmdstanpy:start chain 1
INFO:cmdstanpy:finish chain 1
INFO:cmdstanpy:start chain 1
INFO:cmdstanpy:finish chain 1
     forecast[1]  forecast[2]  forecast[3]
0        21.8522      21.4912      20.8139
1        21.5628      22.9161      22.8557
2        20.1681      19.5121      17.5447
3        20.5012      20.3289      17.0195
4        19.3182      18.3402      17.0739
..           ...          ...          ...
995      21.0422      19.3333      19.0645
996      19.5252      19.5113      17.7110
997      21.4017      21.5539      22.1827
998      19.9526      20.4000      20.5038
999      21.0860      20.8643      25.2736
2

This is I was looking for.

mean_model.sample(iter_sampling=10)

mean_fit = mean_model.sample(data=cmdstanpy_data, output_dir=out_dir, chains=1, iter_sampling=10)