I am currently working on a project conducting measurement invariance analyses for dichotomous using lavaan. I already used similar codes multiple times and have never had problems before. However, now, I get a warning message that no fit measures are available because the model did not coverage (and I need the respective CFI).
Interestingly, I only get this error for weak measurement invariance models (configural) but not for the strict measurement invariance (residual model).
I am comparing the performance in a number series test (correct vs. incorrect) of participants of two cohorts (2009 vs. 2024; therefore, the grouping variable is “year”). Unfortunately, I am not allowed to share the data.
This is the respective code:

numberseriesIST2000 <- '
numberseriesIST2000 =~ zf01_2000_dich+zf02_2000_dich+zf03_2000_dich+zf04_2000_dich+zf05_2000_dich+zf06_2000_dich+zf07_2000_dich+zf08_2000_dich+zf09_2000_dich+zf10_2000_dich+zf11_2000_dich+zf12_2000_dich+zf13_2000_dich+zf14_2000_dich+zf15_2000_dich+zf16_2000_dich+zf17_2000_dich+zf18_2000_dich+zf19_2000_dich+zf20_2000_dich
'

config_model.fitN2000 <- cfa(numberseriesIST2000,
                           data = data,
                           ordered = TRUE,
                           meanstructure = TRUE,
                           group = "Year",
                           group.equal=c("thresholds","loadings"),
                           parameterization = "theta",
                           estimator = "wlsmv",
                           std.lv = TRUE)
summary(config_model.fitN2000, standardized = TRUE, fit.measures=TRUE)

Warning message:
In lav_object_summary(object = object, header = header, fit.measures = fit.measures, :
lavaan WARNING: fit measures not available if model did not converge

residual_model.fitN2000 <- cfa(numberseriesIST2000,
                            data = data,
                            ordered = TRUE,
                            meanstructure = TRUE,
                            group = "Year",
                            group.equal=c("thresholds","loadings","residuals"),
                            parameterization = "theta",
                            estimator = "wlsmv",
                            std.lv = TRUE)
summary(residual_model.fitN2000, standardized = TRUE, fit.measures=TRUE)

I do not get why the model does not have coverage.