predict_response causes "R Session Aborted" with glmer model

[coanv]

Hello!

I am analyzing longitudinal panel data (the health and retirement study) using a mixed effects binary logistic regression model (glmer) with year and person random effects. When I try to use predict_response() I encounter a fatal error that terminates my R session. I am able to successfully run ggeffect(), but would like to understand why predict_response is causing this error.

Here is the model specification I am using
model <- glmer(haveccd ~ cdcombanly + gender + race*age_scl+race*I(age_scl^2)+coll+rmstat_c+employ+homeown+selfhlth+hhhres+loghinc_real+year2+(1|HHIDPN)+ (year2-1|HHIDPN) ,data=githubdta, family=binomial, control =glmerControl(optimizer="nlminbwrap", optCtrl = list(maxfun = 10000000)))

this causes my R session to abort
prdct<- predict_response(CH2_MCNF3,"cdcombanly",bias_correction = TRUE)

this runs without issue
prdct1 <- ggeffect(CH2_MCNF3,c("cdcombanly"), bias_correction = TRUE)

While this model includes quadratic interaction terms, I have encountered the same issue when running a model without age squared and with no interaction between race and age.

I have attached a csv of the data I am using.

githubdta.csv