autogluon.tabular 1.2.1b20250218__tar.gz → 1.2.1b20250219__tar.gz

{autogluon.tabular-1.2.1b20250218 → autogluon.tabular-1.2.1b20250219}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: autogluon.tabular
-Version: 1.2.1b20250218
+Version: 1.2.1b20250219
 Summary: Fast and Accurate ML in 3 Lines of Code
 Home-page: https://github.com/autogluon/autogluon
 Author: AutoGluon Community

{autogluon.tabular-1.2.1b20250218 → autogluon.tabular-1.2.1b20250219}/src/autogluon/tabular/predictor/predictor.py

@@ -3793,6 +3793,80 @@ class TabularPredictor(TabularPredictorDeprecatedMixin):
         self._assert_is_fit("info")
         return self._learner.get_info(include_model_info=True, include_model_failures=True)
 
+    def model_info(self, model: str) -> dict:
+        """
+        Returns metadata information about the given model.
+        Equivalent output to `predictor.info()["model_info"][model]`
+
+        Parameters
+        ----------
+        model: str
+            The name of the model to get info for.
+
+        Returns
+        -------
+        model_info: dict
+            Model info dictionary
+
+        """
+        return self._trainer.get_model_info(model=model)
+
+    # TODO: Add entire `hyperparameters` dict method for multiple models (including stack ensemble)
+    # TODO: Add unit test
+    def model_hyperparameters(
+        self,
+        model: str,
+        include_ag_args_ensemble: bool = True,
+        output_format: Literal["user", "all"] = "user",
+    ) -> dict:
+        """
+        Returns the hyperparameters of a given model.
+
+        Parameters
+        ----------
+        model: str
+            The name of the model to get hyperparameters for.
+        include_ag_args_ensemble: bool, default True
+            If True, includes the ag_args_ensemble parameters if they exist (for example, when bagging is enabled).
+        output_format: {"user", "all"}, default "user"
+            If "user", returns the same hyperparameters specified by the user (only non-defaults).
+            If "all", returns all hyperparameters used by the model (including default hyperparameters not specified by the user)
+            Regardless of the output_format, they both are functionally equivalent if passed to AutoGluon.
+
+        Returns
+        -------
+        model_hyperparameters: dict
+            Dictionary of model hyperparameters.
+            Equivalent to the model_hyperparameters specified by the user for this model in:
+                `predictor.fit(..., hyperparameters={..., model_key: [..., model_hyperparameters]})`
+
+        """
+        # TODO: Move logic into trainer?
+        info_model = self.model_info(model=model)
+        if output_format == "user":
+            if "bagged_info" in info_model:
+                hyperparameters = info_model["bagged_info"]["child_hyperparameters_user"].copy()
+                if include_ag_args_ensemble and info_model["hyperparameters_user"]:
+                    hyperparameters["ag_args_ensemble"] = info_model["hyperparameters_user"]
+            else:
+                hyperparameters = info_model["hyperparameters_user"]
+        elif output_format == "all":
+            if "bagged_info" in info_model:
+                hyperparameters = info_model["bagged_info"]["child_hyperparameters"].copy()
+                if info_model["bagged_info"]["child_ag_args_fit"]:
+                    hyperparameters["ag_args_fit"] = info_model["bagged_info"]["child_ag_args_fit"]
+                if include_ag_args_ensemble:
+                    bag_hyperparameters = info_model["hyperparameters"].copy()
+                    if info_model["ag_args_fit"]:
+                        bag_hyperparameters["ag_args_fit"] = info_model["ag_args_fit"]
+                    if bag_hyperparameters:
+                        hyperparameters["ag_args_ensemble"] = bag_hyperparameters
+            else:
+                hyperparameters = info_model["hyperparameters"]
+        else:
+            raise ValueError(f"output_format={output_format} is unknown!")
+        return hyperparameters
+
     # TODO: Add data argument
     # TODO: Add option to disable OOF generation of newly fitted models
     # TODO: Move code logic to learner/trainer

{autogluon.tabular-1.2.1b20250218 → autogluon.tabular-1.2.1b20250219}/src/autogluon/tabular/version.py

@@ -1,4 +1,4 @@
 """This is the autogluon version file."""
 
-__version__ = "1.2.1b20250218"
+__version__ = "1.2.1b20250219"
 __lite__ = False

{autogluon.tabular-1.2.1b20250218 → autogluon.tabular-1.2.1b20250219}/src/autogluon.tabular.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: autogluon.tabular
-Version: 1.2.1b20250218
+Version: 1.2.1b20250219
 Summary: Fast and Accurate ML in 3 Lines of Code
 Home-page: https://github.com/autogluon/autogluon
 Author: AutoGluon Community

{autogluon.tabular-1.2.1b20250218 → autogluon.tabular-1.2.1b20250219}/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==1.2.1b20250218
-autogluon.features==1.2.1b20250218
+autogluon.core==1.2.1b20250219
+autogluon.features==1.2.1b20250219
 
 [all]
-huggingface_hub[torch]
-xgboost<2.2,>=1.6
-lightgbm<4.6,>=4.0
-spacy<3.8
-autogluon.core[all]==1.2.1b20250218
-numpy<2.0.0,>=1.25
-einops<0.9,>=0.7
 catboost<1.3,>=1.2
 torch<2.6,>=2.2
+spacy<3.8
+lightgbm<4.6,>=4.0
 fastai<2.8,>=2.3.1
+huggingface_hub[torch]
+numpy<2.0.0,>=1.25
+xgboost<2.2,>=1.6
+einops<0.9,>=0.7
+autogluon.core[all]==1.2.1b20250219
 
 [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]==1.2.1b20250218
+autogluon.core[all]==1.2.1b20250219
 
 [skex]
 scikit-learn-intelex<2025.1,>=2024.0