create_forecast_loader

X_forecast, y_forecast = feature_label_split(df_forecast, 'value')

scaler = get_scaler('minmax')
X_train_arr = scaler.fit_transform(X_train)
X_forecast_arr = scaler.transform(X_forecast)
y_train_arr = scaler.fit_transform(y_train)
y_forecast_arr = scaler.transform(y_forecast)

forecast_dataset = TensorDataset(torch.Tensor(X_forecast_arr), 
                                 torch.Tensor(y_forecast_arr))

forecast_loader = DataLoader(
    forecast_dataset, batch_size=1, shuffle=False, drop_last=True
)