Prophet Model method taking longer time to run when we do model.add_seasonality(name='yearly', period=365.25, fourier_order=20)

[Suraj-7077]

Trying to create a forecasting model with having sales data of 26k with parameters as provided.
its taking much longer time to run when we add seasonality name yearly with fourier order =20 and this parameters is the best parameter that are giving me good accuracy with better sales prediction on Holidays and on Sales days.
Note here holidays is data frame of Sales days it has columns like [sales_day_name, ds, lower_window, upper_window. ]

model = Prophet(
           growth='linear',
           holidays=holidays,
           seasonality_mode='multiplicative',
           changepoint_prior_scale=0.05,
           seasonality_prior_scale=35,
           holidays_prior_scale=20,
           interval_width=0.75,
           daily_seasonality=False, # set False bcz not to take auto value since found best value through best_param method(and found fourier order = 12)
           weekly_seasonality=False,  # set False bcz not to take auto value since found best value through best_param method(and found fourier order = 20)
           yearly_seasonality=False,   # set False bcz not to take auto value since found best value through best_param method(and found fourier order = 20)
       ).add_seasonality(name='monthly', period=30.5, fourier_order=12)
       model.add_seasonality(name='weekly', period=7, fourier_order=20)
       model.add_seasonality(name='yearly', period=365.25, fourier_order=20)

And when I remove the yearly seasonality it will runs very fast but that decreases the accuracy of model sales prediction.

How can I reduce the run time without decreasing the Sales prediction accuracy?