Unexpected keyword argument passed to TabPFNRegressor by default

[TheNamesAlex]

First off, thanks a lot for the amazing work you've been doing here. I am trying to play around with the package, and ran into the error below when running the code from your colab demo locally.

A TypeError is thrown when running the snipped below:
TypeError: TabPFNRegressor.init() got an unexpected keyword argument 'optimize_metric'

Expected behavior:
default params match expected constructor args

tabpfn versions used:
tabpfn==2.0.6
tabpfn-client==0.0.25
tabpfn_time_series==0.1.1

snipped to reproduce:
`from datasets import load_dataset
from autogluon.timeseries import TimeSeriesDataFrame

from tabpfn_time_series import FeatureTransformer, DefaultFeatures
from tabpfn_time_series.data_preparation import to_gluonts_univariate, generate_test_X
from tabpfn_time_series.plot import plot_actual_ts
from tabpfn_time_series import TabPFNTimeSeriesPredictor, TabPFNMode

dataset_metadata = {
"monash_tourism_monthly": {"prediction_length": 24},
"m4_hourly": {"prediction_length": 48},
}

dataset_choice = "monash_tourism_monthly"
num_time_series_subset = 2

prediction_length = dataset_metadata[dataset_choice]['prediction_length']
dataset = load_dataset("autogluon/chronos_datasets", dataset_choice)

tsdf = TimeSeriesDataFrame(to_gluonts_univariate(dataset['train']))
tsdf = tsdf[tsdf.index.get_level_values('item_id').isin(tsdf.item_ids[:num_time_series_subset])]
train_tsdf, test_tsdf_ground_truth = tsdf.train_test_split(prediction_length=prediction_length)
test_tsdf = generate_test_X(train_tsdf, prediction_length)

plot_actual_ts(train_tsdf, test_tsdf_ground_truth)

selected_features = [
DefaultFeatures.add_running_index,
DefaultFeatures.add_calendar_features,
]

train_tsdf, test_tsdf = FeatureTransformer.add_features(
train_tsdf, test_tsdf, selected_features
)
train_tsdf.head()

predictor = TabPFNTimeSeriesPredictor(
tabpfn_mode=TabPFNMode.LOCAL
)

pred = predictor.predict(train_tsdf, test_tsdf)`

[liam-sbhoo]

@TheNamesAlex

Thanks for being interested!

Could you try again with tabpfn-time-series==0.1.2? That should probably fix the issue.

[liam-sbhoo]

@TheNamesAlex

Does this solve the issue for you? Let me know :)

[TheNamesAlex]

Hi @liam-sbhoo, thank you for the swift response. That solved the issue for me!

Thanks, and all the best!