autogluon.tabular 1.2.1b20250113__tar.gz → 1.2.1b20250115__tar.gz

{autogluon.tabular-1.2.1b20250113 → autogluon.tabular-1.2.1b20250115}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: autogluon.tabular
-Version: 1.2.1b20250113
+Version: 1.2.1b20250115
 Summary: Fast and Accurate ML in 3 Lines of Code
 Home-page: https://github.com/autogluon/autogluon
 Author: AutoGluon Community

{autogluon.tabular-1.2.1b20250113 → autogluon.tabular-1.2.1b20250115}/setup.py

@@ -52,7 +52,8 @@ extras_require = {
         "fastai>=2.3.1,<2.8",  # <{N+1} upper cap, where N is the latest released minor version
     ],
     "tabpfn": [
-        "tabpfn>=0.1,<0.2",  # <{N+1} upper cap, where N is the latest released minor version
+        # versions below 0.1.11 are broken
+        "tabpfn>=0.1.11,<0.2",  # <{N+1} upper cap, where N is the latest released minor version
     ],
     "tabpfnmix": [
         "torch",  # version range defined in `core/_setup_utils.py`

{autogluon.tabular-1.2.1b20250113 → autogluon.tabular-1.2.1b20250115}/src/autogluon/tabular/models/tabular_nn/torch/tabular_nn_torch.py

@@ -156,12 +156,26 @@ class TabularNeuralNetTorchModel(AbstractNeuralNetworkModel):
 
         return processor_kwargs, optimizer_kwargs, fit_kwargs, loss_kwargs, params
 
-    def _fit(self, X, y, X_val=None, y_val=None, time_limit=None, sample_weight=None, num_cpus=1, num_gpus=0, reporter=None, verbosity=2, **kwargs):
+    def _fit(
+        self,
+        X: pd.DataFrame,
+        y: pd.Series,
+        X_val: pd.DataFrame = None,
+        y_val: pd.Series = None,
+        X_test: pd.DataFrame = None,
+        y_test: pd.Series = None,
+        time_limit: float = None,
+        sample_weight=None,
+        num_cpus: int = 1,
+        num_gpus: float = 0,
+        reporter=None,
+        verbosity: int = 2,
+        **kwargs,
+    ):
         try_import_torch()
         import torch
 
         torch.set_num_threads(num_cpus)
-        from .tabular_torch_dataset import TabularTorchDataset
 
         start_time = time.time()
 
@@ -188,19 +202,20 @@ class TabularNeuralNetTorchModel(AbstractNeuralNetworkModel):
             self.num_dataloading_workers = 0  # TODO: verify 0 is typically faster and uses less memory than 1 in pytorch
         self.num_dataloading_workers = 0  # TODO: >0 crashes on MacOS
         self.max_batch_size = params.pop("max_batch_size", 512)
-        batch_size = params.pop("batch_size", None)
-        if batch_size is None:
-            if isinstance(X, TabularTorchDataset):
-                batch_size = min(int(2 ** (3 + np.floor(np.log10(len(X))))), self.max_batch_size)
-            else:
-                batch_size = min(int(2 ** (3 + np.floor(np.log10(X.shape[0])))), self.max_batch_size)
 
-        X_test = kwargs.get("X_test", None)
-        y_test = kwargs.get("y_test", None)
+        train_dataset = self._generate_dataset(X=X, y=y, train_params=processor_kwargs, is_train=True)
+        if X_val is not None and y_val is not None:
+            val_dataset = self._generate_dataset(X=X_val, y=y_val)
+        else:
+            val_dataset = None
+        if X_test is not None and y_test is not None:
+            test_dataset = self._generate_dataset(X=X_test, y=y_test)
+        else:
+            test_dataset = None
 
-        train_dataset = self._generate_dataset(X, y, train_params=processor_kwargs, is_train=True)
-        val_dataset = self._generate_dataset(X_val, y_val)
-        test_dataset = self._generate_dataset(X_test, y_test)
+        batch_size = params.pop("batch_size", None)
+        if batch_size is None:
+            batch_size = min(int(2 ** (3 + np.floor(np.log10(len(X))))), self.max_batch_size, len(X))
 
         logger.log(
             15,
@@ -255,16 +270,16 @@ class TabularNeuralNetTorchModel(AbstractNeuralNetworkModel):
 
     def _train_net(
         self,
-        train_dataset,
-        loss_kwargs,
-        batch_size,
-        num_epochs,
-        epochs_wo_improve,
-        val_dataset=None,
-        test_dataset=None,
-        time_limit=None,
+        train_dataset: TabularTorchDataset,
+        loss_kwargs: dict,
+        batch_size: int,
+        num_epochs: int,
+        epochs_wo_improve: int,
+        val_dataset: TabularTorchDataset = None,
+        test_dataset: TabularTorchDataset = None,
+        time_limit: float = None,
         reporter=None,
-        verbosity=2,
+        verbosity: int = 2,
     ):
         import torch
 
@@ -634,13 +649,13 @@ class TabularNeuralNetTorchModel(AbstractNeuralNetworkModel):
         preds_dataset = np.concatenate(preds_dataset, 0)
         return preds_dataset
 
-    def _generate_dataset(self, X: pd.DataFrame, y: pd.Series, train_params: dict = {}, is_train: bool = False):
+    def _generate_dataset(self, X: pd.DataFrame | TabularTorchDataset, y: pd.Series, train_params: dict = {}, is_train: bool = False) -> TabularTorchDataset:
         """
         Generate TabularTorchDataset from X and y.
 
         Params:
         -------
-        X: pd.DataFrame
+        X: pd.DataFrame | TabularTorchDataset
             The X data.
         y: pd.Series
             The y data.
@@ -676,14 +691,11 @@ class TabularNeuralNetTorchModel(AbstractNeuralNetworkModel):
                     use_ngram_features=use_ngram_features,
                 )
         else:
-            if X is not None:
-                if isinstance(X, TabularTorchDataset):
-                    dataset = X
-                else:
-                    X = self.preprocess(X)
-                    dataset = self._process_test_data(df=X, labels=y)
+            if isinstance(X, TabularTorchDataset):
+                dataset = X
             else:
-                dataset = None
+                X = self.preprocess(X)
+                dataset = self._process_test_data(df=X, labels=y)
 
         return dataset
 

{autogluon.tabular-1.2.1b20250113 → autogluon.tabular-1.2.1b20250115}/src/autogluon/tabular/predictor/predictor.py

@@ -1388,10 +1388,19 @@ class TabularPredictor(TabularPredictorDeprecatedMixin):
             )
         else:
             # Holdout is false, use (repeated) cross-validation
-            is_stratified = self.problem_type in [REGRESSION, QUANTILE, SOFTCLASS]
+            is_stratified = self.problem_type in [BINARY, MULTICLASS]
+            is_binned = self.problem_type in [REGRESSION, QUANTILE]
             self._learner._validate_groups(X=X, X_val=X_val)  # Validate splits before splitting
-            splits = CVSplitter(n_splits=n_folds, n_repeats=n_repeats, groups=self._learner.groups, stratified=is_stratified, random_state=42).split(
-                X=X.drop(self.label, axis=1), y=X[self.label]
+            splits = CVSplitter(
+                n_splits=n_folds,
+                n_repeats=n_repeats,
+                groups=self._learner.groups,
+                stratify=is_stratified,
+                bin=is_binned,
+                random_state=42,
+            ).split(
+                X=X.drop(self.label, axis=1),
+                y=X[self.label]
             )
             n_splits = len(splits)
             logger.info(

{autogluon.tabular-1.2.1b20250113 → autogluon.tabular-1.2.1b20250115}/src/autogluon/tabular/version.py

@@ -1,3 +1,3 @@
 """This is the autogluon version file."""
-__version__ = '1.2.1b20250113'
+__version__ = '1.2.1b20250115'
 __lite__ = False

{autogluon.tabular-1.2.1b20250113 → autogluon.tabular-1.2.1b20250115}/src/autogluon.tabular.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: autogluon.tabular
-Version: 1.2.1b20250113
+Version: 1.2.1b20250115
 Summary: Fast and Accurate ML in 3 Lines of Code
 Home-page: https://github.com/autogluon/autogluon
 Author: AutoGluon Community

{autogluon.tabular-1.2.1b20250113 → autogluon.tabular-1.2.1b20250115}/src/autogluon.tabular.egg-info/requires.txt

@@ -3,20 +3,20 @@ scipy<1.16,>=1.5.4
 pandas<2.3.0,>=2.0.0
 scikit-learn<1.5.3,>=1.4.0
 networkx<4,>=3.0
-autogluon.core==1.2.1b20250113
-autogluon.features==1.2.1b20250113
+autogluon.core==1.2.1b20250115
+autogluon.features==1.2.1b20250115
 
 [all]
-spacy<3.8
-lightgbm<4.6,>=4.0
 torch<2.6,>=2.2
-autogluon.core[all]==1.2.1b20250113
-fastai<2.8,>=2.3.1
-xgboost<2.2,>=1.6
-catboost<1.3,>=1.2
 einops<0.9,>=0.7
+catboost<1.3,>=1.2
 huggingface_hub[torch]
 numpy<2.0.0,>=1.25
+fastai<2.8,>=2.3.1
+spacy<3.8
+autogluon.core[all]==1.2.1b20250115
+xgboost<2.2,>=1.6
+lightgbm<4.6,>=4.0
 
 [catboost]
 numpy<2.0.0,>=1.25
@@ -34,7 +34,7 @@ imodels<1.4.0,>=1.3.10
 lightgbm<4.6,>=4.0
 
 [ray]
-autogluon.core[all]==1.2.1b20250113
+autogluon.core[all]==1.2.1b20250115
 
 [skex]
 scikit-learn-intelex<2025.1,>=2024.0
@@ -51,7 +51,7 @@ onnx<1.18.0,>=1.13.0
 onnx<1.16.2,>=1.13.0
 
 [tabpfn]
-tabpfn<0.2,>=0.1
+tabpfn<0.2,>=0.1.11
 
 [tabpfnmix]
 torch<2.6,>=2.2
@@ -59,7 +59,7 @@ huggingface_hub[torch]
 einops<0.9,>=0.7
 
 [tests]
-tabpfn<0.2,>=0.1
+tabpfn<0.2,>=0.1.11
 torch<2.6,>=2.2
 huggingface_hub[torch]
 einops<0.9,>=0.7