snowflake-ml-python 1.6.0__py3-none-any.whl → 1.6.2__py3-none-any.whl

snowflake/ml/model/_signatures/utils.py

@@ -110,6 +110,15 @@ def huggingface_pipeline_signature_auto_infer(task: str, params: Dict[str, Any])
     # https://huggingface.co/docs/transformers/en/main_classes/pipelines#transformers.ConversationalPipeline
     # Needs to convert to conversation object.
     if task == "conversational":
+        warnings.warn(
+            (
+                "Conversational pipeline is removed from transformers since 4.42.0. "
+                "Support will be removed from snowflake-ml-python soon."
+            ),
+            category=DeprecationWarning,
+            stacklevel=1,
+        )
+
         return core.ModelSignature(
             inputs=[
                 core.FeatureSpec(name="user_inputs", dtype=core.DataType.STRING, shape=(-1,)),

snowflake/ml/model/models/llm.py

@@ -70,7 +70,9 @@ class LLM:
 
             import peft
 
-            peft_config = peft.PeftConfig.from_pretrained(model_id_or_path, **hub_kwargs)  # type: ignore[attr-defined]
+            peft_config = peft.PeftConfig.from_pretrained(  # type: ignore[no-untyped-call, attr-defined]
+                model_id_or_path, **hub_kwargs
+            )
             if peft_config.peft_type != peft.PeftType.LORA:  # type: ignore[attr-defined]
                 raise ValueError("Only LORA is supported.")
             if peft_config.task_type != peft.TaskType.CAUSAL_LM:  # type: ignore[attr-defined]

snowflake/ml/model/type_hints.py

@@ -1,4 +1,5 @@
 # mypy: disable-error-code="import"
+from enum import Enum
 from typing import (
     TYPE_CHECKING,
     Any,
@@ -232,13 +233,12 @@ class BaseModelSaveOption(TypedDict):
     _legacy_save: NotRequired[bool]
     function_type: NotRequired[Literal["FUNCTION", "TABLE_FUNCTION"]]
     method_options: NotRequired[Dict[str, ModelMethodSaveOptions]]
-    include_pip_dependencies: NotRequired[bool]
+    enable_explainability: NotRequired[bool]
 
 
 class CatBoostModelSaveOptions(BaseModelSaveOption):
     target_methods: NotRequired[Sequence[str]]
     cuda_version: NotRequired[str]
-    enable_explainability: NotRequired[bool]
 
 
 class CustomModelSaveOption(BaseModelSaveOption):
@@ -252,12 +252,10 @@ class SKLModelSaveOptions(BaseModelSaveOption):
 class XGBModelSaveOptions(BaseModelSaveOption):
     target_methods: NotRequired[Sequence[str]]
     cuda_version: NotRequired[str]
-    enable_explainability: NotRequired[bool]
 
 
 class LGBMModelSaveOptions(BaseModelSaveOption):
     target_methods: NotRequired[Sequence[str]]
-    enable_explainability: NotRequired[bool]
 
 
 class SNOWModelSaveOptions(BaseModelSaveOption):
@@ -433,3 +431,11 @@ class Deployment(TypedDict):
     signature: core.ModelSignature
     options: Required[DeployOptions]
     details: NotRequired[DeployDetails]
+
+
+class ModelObjective(Enum):
+    UNKNOWN = "unknown"
+    BINARY_CLASSIFICATION = "binary_classification"
+    MULTI_CLASSIFICATION = "multi_classification"
+    REGRESSION = "regression"
+    RANKING = "ranking"

snowflake/ml/modeling/_internal/constants.py

@@ -1 +1,2 @@
 IN_ML_RUNTIME_ENV_VAR = "IN_SPCS_ML_RUNTIME"
+USE_OPTIMIZED_DATA_INGESTOR = "USE_OPTIMIZED_DATA_INGESTOR"

snowflake/ml/modeling/_internal/local_implementations/pandas_handlers.py

@@ -166,10 +166,10 @@ class PandasTransformHandlers:
             SnowflakeMLException: The input column list does not have one of `X` and `X_test`.
         """
         assert hasattr(self.estimator, "score")  # make type checker happy
-        argspec = inspect.getfullargspec(self.estimator.score)
-        if "X" in argspec.args:
+        params = inspect.signature(self.estimator.score).parameters
+        if "X" in params:
             score_args = {"X": self.dataset[input_cols]}
-        elif "X_test" in argspec.args:
+        elif "X_test" in params:
             score_args = {"X_test": self.dataset[input_cols]}
         else:
             raise exceptions.SnowflakeMLException(
@@ -178,10 +178,10 @@ class PandasTransformHandlers:
             )
 
         if len(label_cols) > 0:
-            label_arg_name = "Y" if "Y" in argspec.args else "y"
+            label_arg_name = "Y" if "Y" in params else "y"
             score_args[label_arg_name] = self.dataset[label_cols].squeeze()
 
-        if sample_weight_col is not None and "sample_weight" in argspec.args:
+        if sample_weight_col is not None and "sample_weight" in params:
             score_args["sample_weight"] = self.dataset[sample_weight_col].squeeze()
 
         score = self.estimator.score(**score_args)

snowflake/ml/modeling/_internal/local_implementations/pandas_trainer.py

@@ -43,14 +43,14 @@ class PandasModelTrainer:
             Trained model
         """
         assert hasattr(self.estimator, "fit")  # Keep mypy happy
-        argspec = inspect.getfullargspec(self.estimator.fit)
+        params = inspect.signature(self.estimator.fit).parameters
         args = {"X": self.dataset[self.input_cols]}
 
         if self.label_cols:
-            label_arg_name = "Y" if "Y" in argspec.args else "y"
+            label_arg_name = "Y" if "Y" in params else "y"
             args[label_arg_name] = self.dataset[self.label_cols].squeeze()
 
-        if self.sample_weight_col is not None and "sample_weight" in argspec.args:
+        if self.sample_weight_col is not None and "sample_weight" in params:
             args["sample_weight"] = self.dataset[self.sample_weight_col].squeeze()
 
         return self.estimator.fit(**args)
@@ -59,6 +59,7 @@ class PandasModelTrainer:
         self,
         expected_output_cols_list: List[str],
         drop_input_cols: Optional[bool] = False,
+        example_output_pd_df: Optional[pd.DataFrame] = None,
     ) -> Tuple[pd.DataFrame, object]:
         """Trains the model using specified features and target columns from the dataset.
         This API is different from fit itself because it would also provide the predict
@@ -69,6 +70,8 @@ class PandasModelTrainer:
                 name as a list. Defaults to None.
             drop_input_cols (Optional[bool]): Boolean to determine whether to
                 drop the input columns from the output dataset.
+            example_output_pd_df (Optional[pd.DataFrame]): Example output dataframe
+                This is not used in PandasModelTrainer. It is used in SnowparkModelTrainer.
 
         Returns:
             Tuple[pd.DataFrame, object]: [predicted dataset, estimator]
@@ -108,13 +111,13 @@ class PandasModelTrainer:
         assert hasattr(self.estimator, "fit")  # make type checker happy
         assert hasattr(self.estimator, "fit_transform")  # make type checker happy
 
-        argspec = inspect.getfullargspec(self.estimator.fit)
+        params = inspect.signature(self.estimator.fit).parameters
         args = {"X": self.dataset[self.input_cols]}
         if self.label_cols:
-            label_arg_name = "Y" if "Y" in argspec.args else "y"
+            label_arg_name = "Y" if "Y" in params else "y"
             args[label_arg_name] = self.dataset[self.label_cols].squeeze()
 
-        if self.sample_weight_col is not None and "sample_weight" in argspec.args:
+        if self.sample_weight_col is not None and "sample_weight" in params:
             args["sample_weight"] = self.dataset[self.sample_weight_col].squeeze()
 
         inference_res = self.estimator.fit_transform(**args)

snowflake/ml/modeling/_internal/model_specifications.py

@@ -53,11 +53,13 @@ class SKLearnModelSpecifications(ModelSpecifications):
 
 class XGBoostModelSpecifications(ModelSpecifications):
     def __init__(self) -> None:
+        import sklearn
         import xgboost
 
         imports: List[str] = ["xgboost"]
         pkgDependencies: List[str] = [
             f"numpy=={np.__version__}",
+            f"scikit-learn=={sklearn.__version__}",
             f"xgboost=={xgboost.__version__}",
             f"cloudpickle=={cp.__version__}",
         ]

snowflake/ml/modeling/_internal/model_trainer.py

@@ -20,6 +20,7 @@ class ModelTrainer(Protocol):
         self,
         expected_output_cols_list: List[str],
         drop_input_cols: Optional[bool] = False,
+        example_output_pd_df: Optional[pd.DataFrame] = None,
     ) -> Tuple[Union[DataFrame, pd.DataFrame], object]:
         raise NotImplementedError
 

snowflake/ml/modeling/_internal/snowpark_implementations/distributed_hpo_trainer.py

@@ -495,7 +495,7 @@ class DistributedHPOTrainer(SnowparkModelTrainer):
                     label_arg_name = "Y" if "Y" in argspec.args else "y"
                     args[label_arg_name] = df[label_cols].squeeze()
 
-                if sample_weight_col is not None and "sample_weight" in argspec.args:
+                if sample_weight_col is not None:
                     args["sample_weight"] = df[sample_weight_col].squeeze()
                 return args, estimator, indices, len(df), params_to_evaluate
 
@@ -1061,7 +1061,7 @@ class DistributedHPOTrainer(SnowparkModelTrainer):
                         if label_cols:
                             label_arg_name = "Y" if "Y" in argspec.args else "y"
                             args[label_arg_name] = y
-                        if sample_weight_col is not None and "sample_weight" in argspec.args:
+                        if sample_weight_col is not None:
                             args["sample_weight"] = df[sample_weight_col].squeeze()
                         # estimator.refit = original_refit
                         refit_start_time = time.time()

snowflake/ml/modeling/_internal/snowpark_implementations/snowpark_handlers.py

@@ -318,19 +318,19 @@ class SnowparkTransformHandlers:
             with open(local_score_file_name_path, mode="r+b") as local_score_file_obj:
                 estimator = cp.load(local_score_file_obj)
 
-            argspec = inspect.getfullargspec(estimator.score)
-            if "X" in argspec.args:
+            params = inspect.signature(estimator.score).parameters
+            if "X" in params:
                 args = {"X": df[input_cols]}
-            elif "X_test" in argspec.args:
+            elif "X_test" in params:
                 args = {"X_test": df[input_cols]}
             else:
                 raise RuntimeError("Neither 'X' or 'X_test' exist in argument")
 
             if label_cols:
-                label_arg_name = "Y" if "Y" in argspec.args else "y"
+                label_arg_name = "Y" if "Y" in params else "y"
                 args[label_arg_name] = df[label_cols].squeeze()
 
-            if sample_weight_col is not None and "sample_weight" in argspec.args:
+            if sample_weight_col is not None and "sample_weight" in params:
                 args["sample_weight"] = df[sample_weight_col].squeeze()
 
             result: float = estimator.score(**args)

snowflake/ml/modeling/_internal/snowpark_implementations/snowpark_trainer.py

@@ -35,6 +35,7 @@ cp.register_pickle_by_value(inspect.getmodule(handle_inference_result))
 
 _PROJECT = "ModelDevelopment"
 _ENABLE_ANONYMOUS_SPROC = False
+_ENABLE_TRACER = True
 
 
 class SnowparkModelTrainer:
@@ -119,6 +120,8 @@ class SnowparkModelTrainer:
             A callable that can be registered as a stored procedure.
         """
         imports = model_spec.imports  # In order for the sproc to not resolve this reference in snowflake.ml
+        method_name = "fit"
+        tracer_name = f"snowpark.ml.modeling.{self._class_name.lower()}.{method_name}"
 
         def fit_wrapper_function(
             session: Session,
@@ -138,110 +141,98 @@ class SnowparkModelTrainer:
             for import_name in imports:
                 importlib.import_module(import_name)
 
-            # Execute snowpark queries and obtain the results as pandas dataframe
-            # NB: this implies that the result data must fit into memory.
-            for query in sql_queries[:-1]:
-                _ = session.sql(query).collect(statement_params=statement_params)
-            sp_df = session.sql(sql_queries[-1])
-            df: pd.DataFrame = sp_df.to_pandas(statement_params=statement_params)
-            df.columns = sp_df.columns
+            def fit_and_return_estimator() -> str:
+                """This is a helper function within the sproc to download the data, fit the model, and upload the model.
+
+                Returns:
+                    The name of the file in session's temp stage (temp_stage_name) that contains the serialized model.
+                """
+                # Execute snowpark queries and obtain the results as pandas dataframe
+                # NB: this implies that the result data must fit into memory.
+                for query in sql_queries[:-1]:
+                    _ = session.sql(query).collect(statement_params=statement_params)
+                sp_df = session.sql(sql_queries[-1])
+                df: pd.DataFrame = sp_df.to_pandas(statement_params=statement_params)
+                df.columns = sp_df.columns
+
+                local_transform_file_name = temp_file_utils.get_temp_file_path()
+
+                session.file.get(
+                    stage_location=temp_stage_name,
+                    target_directory=local_transform_file_name,
+                    statement_params=statement_params,
+                )
 
-            local_transform_file_name = temp_file_utils.get_temp_file_path()
+                local_transform_file_path = os.path.join(
+                    local_transform_file_name, os.listdir(local_transform_file_name)[0]
+                )
+                with open(local_transform_file_path, mode="r+b") as local_transform_file_obj:
+                    estimator = cp.load(local_transform_file_obj)
 
-            session.file.get(
-                stage_location=temp_stage_name,
-                target_directory=local_transform_file_name,
-                statement_params=statement_params,
-            )
+                params = inspect.signature(estimator.fit).parameters
+                args = {"X": df[input_cols]}
+                if label_cols:
+                    label_arg_name = "Y" if "Y" in params else "y"
+                    args[label_arg_name] = df[label_cols].squeeze()
 
-            local_transform_file_path = os.path.join(
-                local_transform_file_name, os.listdir(local_transform_file_name)[0]
-            )
-            with open(local_transform_file_path, mode="r+b") as local_transform_file_obj:
-                estimator = cp.load(local_transform_file_obj)
+                if sample_weight_col is not None and "sample_weight" in params:
+                    args["sample_weight"] = df[sample_weight_col].squeeze()
 
-            argspec = inspect.getfullargspec(estimator.fit)
-            args = {"X": df[input_cols]}
-            if label_cols:
-                label_arg_name = "Y" if "Y" in argspec.args else "y"
-                args[label_arg_name] = df[label_cols].squeeze()
+                estimator.fit(**args)
 
-            if sample_weight_col is not None and "sample_weight" in argspec.args:
-                args["sample_weight"] = df[sample_weight_col].squeeze()
+                local_result_file_name = temp_file_utils.get_temp_file_path()
 
-            estimator.fit(**args)
+                with open(local_result_file_name, mode="w+b") as local_result_file_obj:
+                    cp.dump(estimator, local_result_file_obj)
 
-            local_result_file_name = temp_file_utils.get_temp_file_path()
+                session.file.put(
+                    local_file_name=local_result_file_name,
+                    stage_location=temp_stage_name,
+                    auto_compress=False,
+                    overwrite=True,
+                    statement_params=statement_params,
+                )
+                return local_result_file_name
 
-            with open(local_result_file_name, mode="w+b") as local_result_file_obj:
-                cp.dump(estimator, local_result_file_obj)
+            if _ENABLE_TRACER:
 
-            session.file.put(
-                local_file_name=local_result_file_name,
-                stage_location=temp_stage_name,
-                auto_compress=False,
-                overwrite=True,
-                statement_params=statement_params,
-            )
+                # Use opentelemetry to trace the dist and span of the fit operation.
+                # This would allow user to see the trace in the Snowflake UI.
+                from opentelemetry import trace
 
-            # Note: you can add something like  + "|" + str(df) to the return string
-            # to pass debug information to the caller.
-            return str(os.path.basename(local_result_file_name))
+                tracer = trace.get_tracer(tracer_name)
+                with tracer.start_as_current_span("fit"):
+                    local_result_file_name = fit_and_return_estimator()
+                    # Note: you can add something like  + "|" + str(df) to the return string
+                    # to pass debug information to the caller.
+                    return str(os.path.basename(local_result_file_name))
+            else:
+                local_result_file_name = fit_and_return_estimator()
+                return str(os.path.basename(local_result_file_name))
 
         return fit_wrapper_function
 
-    def _get_fit_wrapper_sproc_anonymous(self, statement_params: Dict[str, str]) -> StoredProcedure:
+    def _get_fit_wrapper_sproc(self, statement_params: Dict[str, str], anonymous: bool) -> StoredProcedure:
         model_spec = ModelSpecificationsBuilder.build(model=self.estimator)
-        fit_sproc_name = snowpark_utils.random_name_for_temp_object(snowpark_utils.TempObjectType.PROCEDURE)
-
-        relaxed_dependencies = pkg_version_utils.get_valid_pkg_versions_supported_in_snowflake_conda_channel(
-            pkg_versions=model_spec.pkgDependencies, session=self.session
-        )
-
-        fit_wrapper_sproc = self.session.sproc.register(
-            func=self._build_fit_wrapper_sproc(model_spec=model_spec),
-            is_permanent=False,
-            name=fit_sproc_name,
-            packages=["snowflake-snowpark-python"] + relaxed_dependencies,  # type: ignore[arg-type]
-            replace=True,
-            session=self.session,
-            statement_params=statement_params,
-            anonymous=True,
-            execute_as="caller",
-        )
-
-        return fit_wrapper_sproc
-
-    def _get_fit_wrapper_sproc(self, statement_params: Dict[str, str]) -> StoredProcedure:
-        # If the sproc already exists, don't register.
-        if not hasattr(self.session, "_FIT_WRAPPER_SPROCS"):
-            self.session._FIT_WRAPPER_SPROCS: Dict[str, StoredProcedure] = {}  # type: ignore[attr-defined, misc]
-
-        model_spec = ModelSpecificationsBuilder.build(model=self.estimator)
-        fit_sproc_key = model_spec.__class__.__name__
-        if fit_sproc_key in self.session._FIT_WRAPPER_SPROCS:  # type: ignore[attr-defined]
-            fit_sproc: StoredProcedure = self.session._FIT_WRAPPER_SPROCS[fit_sproc_key]  # type: ignore[attr-defined]
-            return fit_sproc
 
         fit_sproc_name = snowpark_utils.random_name_for_temp_object(snowpark_utils.TempObjectType.PROCEDURE)
 
         relaxed_dependencies = pkg_version_utils.get_valid_pkg_versions_supported_in_snowflake_conda_channel(
             pkg_versions=model_spec.pkgDependencies, session=self.session
         )
+        packages = ["snowflake-snowpark-python", "snowflake-telemetry-python"] + relaxed_dependencies
 
         fit_wrapper_sproc = self.session.sproc.register(
             func=self._build_fit_wrapper_sproc(model_spec=model_spec),
             is_permanent=False,
             name=fit_sproc_name,
-            packages=["snowflake-snowpark-python"] + relaxed_dependencies,  # type: ignore[arg-type]
+            packages=packages,  # type: ignore[arg-type]
             replace=True,
             session=self.session,
             statement_params=statement_params,
             execute_as="caller",
+            anonymous=anonymous,
         )
-
-        self.session._FIT_WRAPPER_SPROCS[fit_sproc_key] = fit_wrapper_sproc  # type: ignore[attr-defined]
-
         return fit_wrapper_sproc
 
     def _build_fit_predict_wrapper_sproc(
@@ -333,7 +324,9 @@ class SnowparkModelTrainer:
 
             # write into a temp table in sproc and load the table from outside
             session.write_pandas(
-                fit_predict_result_pd, fit_predict_result_name, auto_create_table=True, table_type="temp"
+                fit_predict_result_pd,
+                fit_predict_result_name,
+                overwrite=True,
             )
 
             # Note: you can add something like  + "|" + str(df) to the return string
@@ -414,13 +407,13 @@ class SnowparkModelTrainer:
             with open(local_transform_file_path, mode="r+b") as local_transform_file_obj:
                 estimator = cp.load(local_transform_file_obj)
 
-            argspec = inspect.getfullargspec(estimator.fit)
+            params = inspect.signature(estimator.fit).parameters
             args = {"X": df[input_cols]}
             if label_cols:
-                label_arg_name = "Y" if "Y" in argspec.args else "y"
+                label_arg_name = "Y" if "Y" in params else "y"
                 args[label_arg_name] = df[label_cols].squeeze()
 
-            if sample_weight_col is not None and "sample_weight" in argspec.args:
+            if sample_weight_col is not None and "sample_weight" in params:
                 args["sample_weight"] = df[sample_weight_col].squeeze()
 
             fit_transform_result = estimator.fit_transform(**args)
@@ -477,7 +470,7 @@ class SnowparkModelTrainer:
 
         return fit_transform_wrapper_function
 
-    def _get_fit_predict_wrapper_sproc_anonymous(self, statement_params: Dict[str, str]) -> StoredProcedure:
+    def _get_fit_predict_wrapper_sproc(self, statement_params: Dict[str, str], anonymous: bool) -> StoredProcedure:
         model_spec = ModelSpecificationsBuilder.build(model=self.estimator)
 
         fit_predict_sproc_name = snowpark_utils.random_name_for_temp_object(snowpark_utils.TempObjectType.PROCEDURE)
@@ -494,82 +487,14 @@ class SnowparkModelTrainer:
             replace=True,
             session=self.session,
             statement_params=statement_params,
-            anonymous=True,
+            anonymous=anonymous,
             execute_as="caller",
         )
 
         return fit_predict_wrapper_sproc
 
-    def _get_fit_predict_wrapper_sproc(self, statement_params: Dict[str, str]) -> StoredProcedure:
-        # If the sproc already exists, don't register.
-        if not hasattr(self.session, "_FIT_WRAPPER_SPROCS"):
-            self.session._FIT_WRAPPER_SPROCS: Dict[str, StoredProcedure] = {}  # type: ignore[attr-defined, misc]
-
-        model_spec = ModelSpecificationsBuilder.build(model=self.estimator)
-        fit_predict_sproc_key = model_spec.__class__.__name__ + "_fit_predict"
-        if fit_predict_sproc_key in self.session._FIT_WRAPPER_SPROCS:  # type: ignore[attr-defined]
-            fit_sproc: StoredProcedure = self.session._FIT_WRAPPER_SPROCS[  # type: ignore[attr-defined]
-                fit_predict_sproc_key
-            ]
-            return fit_sproc
-
-        fit_predict_sproc_name = snowpark_utils.random_name_for_temp_object(snowpark_utils.TempObjectType.PROCEDURE)
-
-        relaxed_dependencies = pkg_version_utils.get_valid_pkg_versions_supported_in_snowflake_conda_channel(
-            pkg_versions=model_spec.pkgDependencies, session=self.session
-        )
-
-        fit_predict_wrapper_sproc = self.session.sproc.register(
-            func=self._build_fit_predict_wrapper_sproc(model_spec=model_spec),
-            is_permanent=False,
-            name=fit_predict_sproc_name,
-            packages=["snowflake-snowpark-python"] + relaxed_dependencies,  # type: ignore[arg-type]
-            replace=True,
-            session=self.session,
-            statement_params=statement_params,
-            execute_as="caller",
-        )
-
-        self.session._FIT_WRAPPER_SPROCS[  # type: ignore[attr-defined]
-            fit_predict_sproc_key
-        ] = fit_predict_wrapper_sproc
-
-        return fit_predict_wrapper_sproc
-
-    def _get_fit_transform_wrapper_sproc_anonymous(self, statement_params: Dict[str, str]) -> StoredProcedure:
-        model_spec = ModelSpecificationsBuilder.build(model=self.estimator)
-
-        fit_transform_sproc_name = snowpark_utils.random_name_for_temp_object(snowpark_utils.TempObjectType.PROCEDURE)
-
-        relaxed_dependencies = pkg_version_utils.get_valid_pkg_versions_supported_in_snowflake_conda_channel(
-            pkg_versions=model_spec.pkgDependencies, session=self.session
-        )
-
-        fit_transform_wrapper_sproc = self.session.sproc.register(
-            func=self._build_fit_transform_wrapper_sproc(model_spec=model_spec),
-            is_permanent=False,
-            name=fit_transform_sproc_name,
-            packages=["snowflake-snowpark-python"] + relaxed_dependencies,  # type: ignore[arg-type]
-            replace=True,
-            session=self.session,
-            statement_params=statement_params,
-            anonymous=True,
-            execute_as="caller",
-        )
-        return fit_transform_wrapper_sproc
-
-    def _get_fit_transform_wrapper_sproc(self, statement_params: Dict[str, str]) -> StoredProcedure:
-        # If the sproc already exists, don't register.
-        if not hasattr(self.session, "_FIT_WRAPPER_SPROCS"):
-            self.session._FIT_WRAPPER_SPROCS: Dict[str, StoredProcedure] = {}  # type: ignore[attr-defined, misc]
-
+    def _get_fit_transform_wrapper_sproc(self, statement_params: Dict[str, str], anonymous: bool) -> StoredProcedure:
         model_spec = ModelSpecificationsBuilder.build(model=self.estimator)
-        fit_transform_sproc_key = model_spec.__class__.__name__ + "_fit_transform"
-        if fit_transform_sproc_key in self.session._FIT_WRAPPER_SPROCS:  # type: ignore[attr-defined]
-            fit_sproc: StoredProcedure = self.session._FIT_WRAPPER_SPROCS[  # type: ignore[attr-defined]
-                fit_transform_sproc_key
-            ]
-            return fit_sproc
 
         fit_transform_sproc_name = snowpark_utils.random_name_for_temp_object(snowpark_utils.TempObjectType.PROCEDURE)
 
@@ -586,12 +511,9 @@ class SnowparkModelTrainer:
             session=self.session,
             statement_params=statement_params,
             execute_as="caller",
+            anonymous=anonymous,
         )
 
-        self.session._FIT_WRAPPER_SPROCS[  # type: ignore[attr-defined]
-            fit_transform_sproc_key
-        ] = fit_transform_wrapper_sproc
-
         return fit_transform_wrapper_sproc
 
     def train(self) -> object:
@@ -629,9 +551,9 @@ class SnowparkModelTrainer:
         # Call fit sproc
 
         if _ENABLE_ANONYMOUS_SPROC:
-            fit_wrapper_sproc = self._get_fit_wrapper_sproc_anonymous(statement_params=statement_params)
+            fit_wrapper_sproc = self._get_fit_wrapper_sproc(statement_params=statement_params, anonymous=True)
         else:
-            fit_wrapper_sproc = self._get_fit_wrapper_sproc(statement_params=statement_params)
+            fit_wrapper_sproc = self._get_fit_wrapper_sproc(statement_params=statement_params, anonymous=False)
 
         try:
             sproc_export_file_name: str = fit_wrapper_sproc(
@@ -665,6 +587,7 @@ class SnowparkModelTrainer:
         self,
         expected_output_cols_list: List[str],
         drop_input_cols: Optional[bool] = False,
+        example_output_pd_df: Optional[pd.DataFrame] = None,
     ) -> Tuple[Union[DataFrame, pd.DataFrame], object]:
         """Trains the model by pushing down the compute into Snowflake using stored procedures.
         This API is different from fit itself because it would also provide the predict
@@ -675,6 +598,11 @@ class SnowparkModelTrainer:
                 name as a list. Defaults to None.
             drop_input_cols (Optional[bool]): Boolean to determine drop
                 the input columns from the output dataset or not
+            example_output_pd_df (Optional[pd.DataFrame]): Example output dataframe
+                This is to create a temp table in the client side with df_one_row. This can maintain the same column
+                name and data type as the output dataframe. Within the sproc, we don't need to create another temp table
+                again - instead, we overwrite into this table without changing the schema.
+                This is not used in PandasModelTrainer.
 
         Returns:
             Tuple[Union[DataFrame, pd.DataFrame], object]: [predicted dataset, estimator]
@@ -702,12 +630,35 @@ class SnowparkModelTrainer:
 
         # Call fit sproc
         if _ENABLE_ANONYMOUS_SPROC:
-            fit_predict_wrapper_sproc = self._get_fit_predict_wrapper_sproc_anonymous(statement_params=statement_params)
+            fit_predict_wrapper_sproc = self._get_fit_predict_wrapper_sproc(
+                statement_params=statement_params, anonymous=True
+            )
         else:
-            fit_predict_wrapper_sproc = self._get_fit_predict_wrapper_sproc(statement_params=statement_params)
+            fit_predict_wrapper_sproc = self._get_fit_predict_wrapper_sproc(
+                statement_params=statement_params, anonymous=False
+            )
 
         fit_predict_result_name = snowpark_utils.random_name_for_temp_object(snowpark_utils.TempObjectType.TABLE)
 
+        # Create a temp table in advance to store the output
+        # This would allow us to use the same table outside the stored procedure
+        if not drop_input_cols:
+            assert example_output_pd_df is not None
+            remove_dataset_col_name_exist_in_output_col = list(set(dataset.columns) - set(example_output_pd_df.columns))
+            pd_df_one_row = (
+                dataset.select(remove_dataset_col_name_exist_in_output_col)
+                .limit(1)
+                .to_pandas(statement_params=statement_params)
+            )
+            example_output_pd_df = pd.concat([pd_df_one_row, example_output_pd_df], axis=1)
+
+        self.session.write_pandas(
+            example_output_pd_df,
+            fit_predict_result_name,
+            auto_create_table=True,
+            table_type="temp",
+        )
+
         sproc_export_file_name: str = fit_predict_wrapper_sproc(
             self.session,
             queries,
@@ -769,11 +720,13 @@ class SnowparkModelTrainer:
 
         # Call fit sproc
         if _ENABLE_ANONYMOUS_SPROC:
-            fit_transform_wrapper_sproc = self._get_fit_transform_wrapper_sproc_anonymous(
-                statement_params=statement_params
+            fit_transform_wrapper_sproc = self._get_fit_transform_wrapper_sproc(
+                statement_params=statement_params, anonymous=True
             )
         else:
-            fit_transform_wrapper_sproc = self._get_fit_transform_wrapper_sproc(statement_params=statement_params)
+            fit_transform_wrapper_sproc = self._get_fit_transform_wrapper_sproc(
+                statement_params=statement_params, anonymous=False
+            )
 
         fit_transform_result_name = snowpark_utils.random_name_for_temp_object(snowpark_utils.TempObjectType.TABLE)