Discussions

I am trying to produce a 1 year simulation using:

timegpt_fcst_df = nixtla_client.forecast(df=df,
X_df=X_df,
h=1_365_24,
time_col=time_col,
target_col=target_col,
num_partitions=10,
#hist_exog_list=hist_exog_list,
freq='h',
model='timegpt-1-long-horizon',
)
timegpt_fcst_df

where df contains the target variable for hourly-data for y years and 380*3 exogenous variables and I get the error The payload is too large. Set num_partitions=3. I have set different "num_partitions" values and I still get the same error. Any ideas?