[Bug]: pipelines fitting fails on time series classification

[Lopa10ko]

Problem

Getting ValueError: Initial pipeline fit was failed due to: catboost/private/libs/algo/data.cpp:199: Dataset test #0 contains class label "12" that is not present in the learn dataset. Check pipeline structure and the correctness of the data

Steps to reproduce

from fedot_ind.api.main import FedotIndustrial
from fedot_ind.tools.loader import DataLoader

dataset_name = 'Handwriting'
loader = DataLoader(dataset_name=dataset_name)
train_data, test_data = loader.load_data()

api_config = dict(problem='classification',
                  metric='f1',
                  timeout=3,
                  pop_size=10,
                  with_tuning=True,
                  seed=42,
                  n_jobs=-1,
                  logging_level=20)
model = FedotIndustrial(**api_config)
model.fit(train_data)