fix: relax check on dataloader (#106)

* fix: relax check on dataloader

* Update CHANGELOG.rst

Author: xuyxu

CHANGELOG.rst

@@ -18,6 +18,7 @@ Changelog
 Ver 0.1.*
 ---------
 
+* |Fix| Relax check on input dataloader | `@xuyxu <https://github.com/xuyxu>`__
 * |Feature| |API| Support arbitrary training criteria for all ensembles except Gradient Boosting | `@by256 <https://github.com/by256>`__ and `@xuyxu <https://github.com/xuyxu>`__
 * |Fix| Fix missing functionality of ``save_model`` for :meth:`fit` of Soft Gradient Boosting | `@xuyxu <https://github.com/xuyxu>`__
 * |Feature| |API| Add :class:`SoftGradientBoostingClassifier` and :class:`SoftGradientBoostingRegressor` | `@xuyxu <https://github.com/xuyxu>`__

torchensemble/utils/io.py

@@ -75,7 +75,7 @@ def load(model, save_dir="./", logger=None):
 
 def split_data_target(element, device, logger=None):
     """Split elements in dataloader according to pre-defined rules."""
-    if not isinstance(element, list):
+    if not (isinstance(element, list) or isinstance(element, tuple)):
         msg = (
             "Invalid dataloader, please check if the input dataloder is valid."
         )