How to set initial value for variable without regression formula in a non-hierarchical model?

[SaschaFroelich]

Let's say I want to initialize a with a value of 0.7. In this thread, When I initialize like this:

model_reg_v_angle = hssm.HSSM(
    data = ddmdata_onesub,
    model = "angle",
    hierarchical = False,
    categorical = 'jokercondition',
    a = bmb.Prior("Uniform", lower=0.1, upper=2.0, initval=0.8),
    include=[
        {
            "name": "v",
            "formula": "v ~ 1 + jokercondition",
            "prior": {
                "Intercept": {"name": "Normal", 
                              "mu": 1, 
                              "sigma": 2, 
                              "initval": initvalues['v_Intercept']},
                "jokercondition": {"name": "Normal", 
                                   "mu": 0, 
                                   "sigma": 2, 
                                   "initval": initvalues['v_jokercondition']},
            },
            "link": "identity",
        },
    ],
)

The initial value for a does not take effect, while the initial values set in include = [... do. Prof. Frank also said in this thread that I shouldn't initialize a like I did above but instead do it in include = [..., however that was for a hierarchical model, where I could specify a regression formula for alike so:

"formula": "a ~ 1 + (1|participant_id)"

Now in the present case however, I would like to fit only one individual participant and therefore, specify a without regressors. I have tried for example "formula": "a ~ 1", but this throws the error AttributeError: 'Intercept' object has no attribute 'eval'. Simply omitting formula throws the error TypeError: Prior.__init__() missing 1 required positional argument: 'name'

How can I initialize a in a flat model?