`PopularInCategoryModel` fails to fit second time

[blondered]

What happened?

    category_feature='genre',
    n_categories=5,
    popularity="mean_weight",
    mixing_strategy="group",
    ratio_strategy="proportional",)
pop_cat.fit(dataset)
pop_cat.fit(dataset)``` fails with TypeError

`cross_validate` fails with the same error

### Expected behavior

No errors :)

### Relevant logs and/or screenshots

_No response_

### Operating System

macOS Big Sur 11.6 (Apple M1)

### Python Version

any

### RecTools version

0.6.0