Python Statsmodels quasi-information criteria qic. Any example?

I must be doing something really wrong, since I simply cannot write a code to obtain the quasi-information criteria of the Python statsmodels GEE method.
This method appears in the same way as the "fit" method used in the code below:

model2 = sm.GEE.from_formula("BPXSY1 ~ RIDAGEYR + RIAGENDRx + BMXBMI + C(RIDRETH1)",groups="group",cov_struct=sm.cov_struct.Exchangeable(), data=da)
result2 =

But if I code "qic" in the same way that I encoded "fit", the script does not run and the following message appears:

AttributeError                            Traceback (most recent call last)
 in ()
----> 1 criteria = model2.qic()

AttributeError: 'GEE' object has no attribute 'qic'

I have also tried:

criteria = model2.qic(result2.params)


criteria = result2.qic()

Nothing works. I always receive the same error message. It must be a syntax error. Can anyone help?
Thank you