[BUG] After calling fbporphet model.fit(), no forecasts all null

[PeggyFan]

Describe the bug
I ran the fbprophet model

clf = HTSRegressor(model='prophet', revision_method='AHP', n_jobs=5)
model = clf.fit(df, hierarchy, exogenous=exogenous)
model.hts_result.forecasts

but it returned:
{}

So naturally all the predictions returned nan.

I have successfully ran a two level model, but not sure why three level hierarchy here isn't working.

To Reproduce
Steps to reproduce the behavior:
df:

reporting_day | total | Flowers | Flowers-01 | Flowers-02 | flowers_promo_pct
2021-05-01 | 36517.00000 | 36517.00000 | 31696.00000 | 4821.00000 | 0.03000
2021-05-02 | 80410.00000 | 80410.00000 | 52164.00000 | 28246.00000 | 0.92000
2021-05-03 | 35210.00000 | 35210.00000 | 28169.00000 | 7041.00000 | 0.02000
2021-05-04 | 52560.00000 | 52560.00000 | 36546.00000 | 16014.00000 | 0.02000
2021-05-05 | 37549.00000 | 37549.00000 | 32744.00000 | 4805.00000 | 0.73000

hierarchy:

{'total': ['Flowers'],
 'Flowers': ['Flowers-01',
  'Flowers-01']}

exogenous:

{'Flowers-01': ['flowers_promo_pct'],
 'Flowers-02': ['flowers_promo_pct'],
 'Flowers': ['flowers_promo_pct'],
 'total': ['flowers_promo_pct']}

Expected behavior
I believe the inputs to the model.fit() are in the correct format. So after I called fit() there should be forecasts under
model.hts_result.forecasts

Desktop (please complete the following information):

• OS: OS X Catalina
• scikit-hts version: 0.5.11
• Python version: 3.8.6

Additional context
Anyone who can spot something wrong with my code, please do share your insights!