snowflake-ml-python 1.3.1__py3-none-any.whl → 1.4.1__py3-none-any.whl

snowflake/ml/model/_packager/model_handlers/sklearn.py

@@ -47,6 +47,9 @@ class SKLModelHandler(_base.BaseModelHandler[Union["sklearn.base.BaseEstimator",
                 or type_utils.LazyType("sklearn.pipeline.Pipeline").isinstance(model)
             )
             and (not type_utils.LazyType("xgboost.XGBModel").isinstance(model))  # XGBModel is actually a BaseEstimator
+            and (
+                not type_utils.LazyType("lightgbm.LGBMModel").isinstance(model)
+            )  # LGBMModel is actually a BaseEstimator
             and any(
                 (hasattr(model, method) and callable(getattr(model, method, None)))
                 for method in cls.DEFAULT_TARGET_METHODS

snowflake/ml/model/_packager/model_handlers/snowmlmodel.py

@@ -87,7 +87,19 @@ class SnowMLModelHandler(_base.BaseModelHandler["BaseEstimator"]):
                     stacklevel=2,
                 )
             assert hasattr(model, "model_signatures"), "Model does not have model signatures as expected."
-            model_meta.signatures = getattr(model, "model_signatures", {})
+            model_signature_dict = getattr(model, "model_signatures", {})
+            target_methods = kwargs.pop("target_methods", None)
+            if not target_methods:
+                model_meta.signatures = model_signature_dict
+            else:
+                temp_model_signature_dict = {}
+                for method_name in target_methods:
+                    method_model_signature = model_signature_dict.get(method_name, None)
+                    if method_model_signature is not None:
+                        temp_model_signature_dict[method_name] = method_model_signature
+                    else:
+                        raise ValueError(f"Target method {method_name} does not exist in the model.")
+                model_meta.signatures = temp_model_signature_dict
 
         model_blob_path = os.path.join(model_blobs_dir_path, name)
         os.makedirs(model_blob_path, exist_ok=True)

snowflake/ml/model/_packager/model_handlers/xgboost.py

@@ -45,7 +45,7 @@ class XGBModelHandler(_base.BaseModelHandler[Union["xgboost.Booster", "xgboost.X
     _HANDLER_MIGRATOR_PLANS: Dict[str, Type[base_migrator.BaseModelHandlerMigrator]] = {}
 
     MODELE_BLOB_FILE_OR_DIR = "model.ubj"
-    DEFAULT_TARGET_METHODS = ["apply", "predict", "predict_proba"]
+    DEFAULT_TARGET_METHODS = ["predict", "predict_proba"]
 
     @classmethod
     def can_handle(

snowflake/ml/model/_packager/model_meta/_core_requirements.py

@@ -4,7 +4,7 @@ REQUIREMENTS = [
     "cloudpickle>=2.0.0",
     "numpy>=1.23,<2",
     "packaging>=20.9,<24",
-    "pandas>=1.0.0,<2",
+    "pandas>=1.0.0,<3",
     "pyyaml>=6.0,<7",
     "snowflake-snowpark-python>=1.11.1,<2,!=1.12.0",
     "typing-extensions>=4.1.0,<5"

snowflake/ml/model/_packager/model_meta/model_meta.py

@@ -23,6 +23,7 @@ from snowflake.ml.model._packager.model_meta import (
     model_meta_schema,
 )
 from snowflake.ml.model._packager.model_meta_migrator import migrator_plans
+from snowflake.ml.model._packager.model_runtime import model_runtime
 
 MODEL_METADATA_FILE = "model.yaml"
 MODEL_CODE_DIR = "code"
@@ -115,7 +116,6 @@ def create_model_metadata(
         python_version=python_version,
         embed_local_ml_library=embed_local_ml_library,
         legacy_save=legacy_save,
-        relax_version=relax_version,
     )
 
     if embed_local_ml_library:
@@ -156,6 +156,8 @@ def create_model_metadata(
                     cloudpickle.register_pickle_by_value(mod)
                     imported_modules.append(mod)
         yield model_meta
+        if relax_version:
+            model_meta.env.relax_version()
         model_meta.save(model_dir_path)
     finally:
         for mod in imported_modules:
@@ -169,7 +171,6 @@ def _create_env_for_model_metadata(
     python_version: Optional[str] = None,
     embed_local_ml_library: bool = False,
     legacy_save: bool = False,
-    relax_version: bool = False,
 ) -> model_env.ModelEnv:
     env = model_env.ModelEnv()
 
@@ -197,10 +198,6 @@ def _create_env_for_model_metadata(
             ],
             check_local_version=True,
         )
-
-    if relax_version:
-        env.relax_version()
-
     return env
 
 
@@ -237,6 +234,7 @@ class ModelMetadata:
         name: str,
         env: model_env.ModelEnv,
         model_type: model_types.SupportedModelHandlerType,
+        runtimes: Optional[Dict[str, model_runtime.ModelRuntime]] = None,
         signatures: Optional[Dict[str, model_signature.ModelSignature]] = None,
         metadata: Optional[Dict[str, str]] = None,
         creation_timestamp: Optional[str] = None,
@@ -262,6 +260,8 @@ class ModelMetadata:
         if models:
             self.models = models
 
+        self._runtimes = runtimes
+
         self.original_metadata_version = original_metadata_version
 
     @property
@@ -273,6 +273,19 @@ class ModelMetadata:
         parsed_min_snowpark_ml_version = version.parse(min_snowpark_ml_version)
         self._min_snowpark_ml_version = max(self._min_snowpark_ml_version, parsed_min_snowpark_ml_version)
 
+    @property
+    def runtimes(self) -> Dict[str, model_runtime.ModelRuntime]:
+        if self._runtimes and "cpu" in self._runtimes:
+            return self._runtimes
+        runtimes = {
+            "cpu": model_runtime.ModelRuntime("cpu", self.env),
+        }
+        if self.env.cuda_version:
+            runtimes.update(
+                {"gpu": model_runtime.ModelRuntime("gpu", self.env, is_gpu=True, server_availability_source="conda")}
+            )
+        return runtimes
+
     def save(self, model_dir_path: str) -> None:
         """Save the model metadata
 
@@ -291,6 +304,10 @@ class ModelMetadata:
             {
                 "creation_timestamp": self.creation_timestamp,
                 "env": self.env.save_as_dict(pathlib.Path(model_dir_path)),
+                "runtimes": {
+                    runtime_name: runtime.save(pathlib.Path(model_dir_path))
+                    for runtime_name, runtime in self.runtimes.items()
+                },
                 "metadata": self.metadata,
                 "model_type": self.model_type,
                 "models": {model_name: blob.to_dict() for model_name, blob in self.models.items()},
@@ -302,6 +319,7 @@ class ModelMetadata:
         )
 
         with open(model_yaml_path, "w", encoding="utf-8") as out:
+            yaml.SafeDumper.ignore_aliases = lambda *args: True  # type: ignore[method-assign]
             yaml.safe_dump(
                 model_dict,
                 stream=out,
@@ -330,6 +348,7 @@ class ModelMetadata:
         return model_meta_schema.ModelMetadataDict(
             creation_timestamp=loaded_meta["creation_timestamp"],
             env=loaded_meta["env"],
+            runtimes=loaded_meta.get("runtimes", None),
             metadata=loaded_meta.get("metadata", None),
             model_type=loaded_meta["model_type"],
             models=loaded_meta["models"],
@@ -363,10 +382,21 @@ class ModelMetadata:
         models = {name: model_blob_meta.ModelBlobMeta(**blob_meta) for name, blob_meta in model_dict["models"].items()}
         env = model_env.ModelEnv()
         env.load_from_dict(pathlib.Path(model_dir_path), model_dict["env"])
+
+        runtimes: Optional[Dict[str, model_runtime.ModelRuntime]]
+        if model_dict.get("runtimes", None):
+            runtimes = {
+                name: model_runtime.ModelRuntime.load(pathlib.Path(model_dir_path), name, env, runtime_dict)
+                for name, runtime_dict in model_dict["runtimes"].items()
+            }
+        else:
+            runtimes = None
+
         return cls(
             name=model_dict["name"],
             model_type=model_dict["model_type"],
             env=env,
+            runtimes=runtimes,
             signatures=signatures,
             metadata=model_dict.get("metadata", None),
             creation_timestamp=model_dict["creation_timestamp"],

snowflake/ml/model/_packager/model_meta/model_meta_schema.py

@@ -1,7 +1,7 @@
 # This files contains schema definition of what will be written into model.yml
 # Changing this file should lead to a change of the schema version.
 
-from typing import Any, Dict, Optional, TypedDict, Union
+from typing import Any, Dict, List, Optional, TypedDict, Union
 
 from typing_extensions import NotRequired, Required
 
@@ -11,6 +11,16 @@ MODEL_METADATA_VERSION = "2023-12-01"
 MODEL_METADATA_MIN_SNOWPARK_ML_VERSION = "1.0.12"
 
 
+class ModelRuntimeDependenciesDict(TypedDict):
+    conda: Required[str]
+    pip: Required[str]
+
+
+class ModelRuntimeDict(TypedDict):
+    imports: Required[List[str]]
+    dependencies: Required[ModelRuntimeDependenciesDict]
+
+
 class ModelEnvDict(TypedDict):
     conda: Required[str]
     pip: Required[str]
@@ -23,11 +33,19 @@ class BaseModelBlobOptions(TypedDict):
     ...
 
 
+class CatBoostModelBlobOptions(BaseModelBlobOptions):
+    catboost_estimator_type: Required[str]
+
+
 class HuggingFacePipelineModelBlobOptions(BaseModelBlobOptions):
     task: Required[str]
     batch_size: Required[int]
 
 
+class LightGBMModelBlobOptions(BaseModelBlobOptions):
+    lightgbm_estimator_type: Required[str]
+
+
 class LLMModelBlobOptions(BaseModelBlobOptions):
     batch_size: Required[int]
 
@@ -61,6 +79,7 @@ class ModelBlobMetadataDict(TypedDict):
 class ModelMetadataDict(TypedDict):
     creation_timestamp: Required[str]
     env: Required[ModelEnvDict]
+    runtimes: NotRequired[Dict[str, ModelRuntimeDict]]
     metadata: NotRequired[Optional[Dict[str, str]]]
     model_type: Required[type_hints.SupportedModelHandlerType]
     models: Required[Dict[str, ModelBlobMetadataDict]]

snowflake/ml/model/_packager/model_meta_migrator/migrator_plans.py

@@ -3,7 +3,9 @@ from typing import Any, Dict, Type
 from snowflake.ml.model._packager.model_meta import model_meta_schema
 from snowflake.ml.model._packager.model_meta_migrator import base_migrator, migrator_v1
 
-MODEL_META_MIGRATOR_PLANS: Dict[str, Type[base_migrator.BaseModelMetaMigrator]] = {"1": migrator_v1.MetaMigrator_v1}
+MODEL_META_MIGRATOR_PLANS: Dict[str, Type[base_migrator.BaseModelMetaMigrator]] = {
+    "1": migrator_v1.MetaMigrator_v1,
+}
 
 
 def migrate_metadata(loaded_meta: Dict[str, Any]) -> Dict[str, Any]:

snowflake/ml/model/_packager/model_packager.py

@@ -102,8 +102,8 @@ class ModelPackager:
             if signatures is None:
                 logging.info(f"Model signatures are auto inferred as:\n\n{meta.signatures}")
 
-            self.model = model
-            self.meta = meta
+        self.model = model
+        self.meta = meta
 
     def load(
         self,

snowflake/ml/model/{_model_composer/model_runtime/_runtime_requirements.py → _packager/model_runtime/_snowml_inference_alternative_requirements.py}

@@ -3,7 +3,7 @@ REQUIREMENTS = [
     "anyio>=3.5.0,<4",
     "numpy>=1.23,<2",
     "packaging>=20.9,<24",
-    "pandas>=1.0.0,<2",
+    "pandas>=1.0.0,<3",
     "pyyaml>=6.0,<7",
     "snowflake-snowpark-python>=1.11.1,<2,!=1.12.0",
     "typing-extensions>=4.1.0,<5"

snowflake/ml/model/_packager/model_runtime/model_runtime.py

@@ -0,0 +1,137 @@
+import copy
+import pathlib
+import warnings
+from typing import List, Literal, Optional
+
+from packaging import requirements
+
+from snowflake.ml._internal import env as snowml_env, env_utils, file_utils
+from snowflake.ml.model._packager.model_env import model_env
+from snowflake.ml.model._packager.model_meta import model_meta_schema
+from snowflake.ml.model._packager.model_runtime import (
+    _snowml_inference_alternative_requirements,
+)
+
+_SNOWML_INFERENCE_ALTERNATIVE_DEPENDENCIES = [
+    str(env_utils.get_package_spec_with_supported_ops_only(requirements.Requirement(r)))
+    for r in _snowml_inference_alternative_requirements.REQUIREMENTS
+]
+
+
+class ModelRuntime:
+    """Class to represent runtime in a model, which controls the runtime and version, imports and dependencies.
+
+    Attributes:
+        runtime_env: ModelEnv object representing the actual environment when deploying. The environment is based on
+            the environment from the packaged model with additional dependencies required to deploy.
+        imports: List of files to be imported in the created functions. At least packed model should be imported.
+            If the required Snowpark ML library is not available in the server-side, we will automatically pack the
+            local version as well as "snowflake-ml-python.zip" and added into the imports.
+    """
+
+    RUNTIME_DIR_REL_PATH = "runtimes"
+
+    def __init__(
+        self,
+        name: str,
+        env: model_env.ModelEnv,
+        imports: Optional[List[pathlib.PurePosixPath]] = None,
+        is_gpu: bool = False,
+        server_availability_source: Literal["snowflake", "conda"] = "snowflake",
+        loading_from_file: bool = False,
+    ) -> None:
+        self.name = name
+        self.runtime_env = copy.deepcopy(env)
+        self.imports = imports or []
+
+        if loading_from_file:
+            return
+
+        snowml_pkg_spec = f"{env_utils.SNOWPARK_ML_PKG_NAME}=={self.runtime_env.snowpark_ml_version}"
+        if self.runtime_env._snowpark_ml_version.local:
+            self.embed_local_ml_library = True
+        else:
+            if server_availability_source == "snowflake":
+                snowml_server_availability = (
+                    len(
+                        env_utils.get_matched_package_versions_in_information_schema_with_active_session(
+                            reqs=[requirements.Requirement(snowml_pkg_spec)],
+                            python_version=snowml_env.PYTHON_VERSION,
+                        ).get(env_utils.SNOWPARK_ML_PKG_NAME, [])
+                    )
+                    >= 1
+                )
+            else:
+                snowml_server_availability = (
+                    len(
+                        env_utils.get_matched_package_versions_in_snowflake_conda_channel(
+                            req=requirements.Requirement(snowml_pkg_spec),
+                            python_version=snowml_env.PYTHON_VERSION,
+                        )
+                    )
+                    >= 1
+                )
+            self.embed_local_ml_library = not snowml_server_availability
+
+        additional_package = (
+            _SNOWML_INFERENCE_ALTERNATIVE_DEPENDENCIES if self.embed_local_ml_library else [snowml_pkg_spec]
+        )
+
+        self.runtime_env.include_if_absent(
+            [
+                model_env.ModelDependency(requirement=dep, pip_name=requirements.Requirement(dep).name)
+                for dep in additional_package
+            ],
+        )
+
+        if is_gpu:
+            self.runtime_env.generate_env_for_cuda()
+
+    @property
+    def runtime_rel_path(self) -> pathlib.PurePosixPath:
+        return pathlib.PurePosixPath(ModelRuntime.RUNTIME_DIR_REL_PATH) / self.name
+
+    def save(self, packager_path: pathlib.Path) -> model_meta_schema.ModelRuntimeDict:
+        runtime_base_path = packager_path / self.runtime_rel_path
+        runtime_base_path.mkdir(parents=True, exist_ok=True)
+
+        if getattr(self, "embed_local_ml_library", False):
+            snowpark_ml_lib_path = runtime_base_path / "snowflake-ml-python.zip"
+            file_utils.zip_python_package(str(snowpark_ml_lib_path), "snowflake.ml")
+            snowpark_ml_lib_rel_path = pathlib.PurePosixPath(snowpark_ml_lib_path.relative_to(packager_path).as_posix())
+            self.imports.append(snowpark_ml_lib_rel_path)
+
+        self.runtime_env.conda_env_rel_path = self.runtime_rel_path / self.runtime_env.conda_env_rel_path
+        self.runtime_env.pip_requirements_rel_path = self.runtime_rel_path / self.runtime_env.pip_requirements_rel_path
+
+        env_dict = self.runtime_env.save_as_dict(packager_path)
+
+        return model_meta_schema.ModelRuntimeDict(
+            imports=list(map(str, self.imports)),
+            dependencies=model_meta_schema.ModelRuntimeDependenciesDict(
+                conda=env_dict["conda"],
+                pip=env_dict["pip"],
+            ),
+        )
+
+    @staticmethod
+    def load(
+        packager_path: pathlib.Path,
+        name: str,
+        meta_env: model_env.ModelEnv,
+        loaded_dict: model_meta_schema.ModelRuntimeDict,
+    ) -> "ModelRuntime":
+        env = model_env.ModelEnv()
+        env.python_version = meta_env.python_version
+        env.cuda_version = meta_env.cuda_version
+        env.snowpark_ml_version = meta_env.snowpark_ml_version
+
+        conda_env_rel_path = pathlib.PurePosixPath(loaded_dict["dependencies"]["conda"])
+        pip_requirements_rel_path = pathlib.PurePosixPath(loaded_dict["dependencies"]["pip"])
+        with warnings.catch_warnings():
+            warnings.simplefilter("ignore")
+            env.load_from_conda_file(packager_path / conda_env_rel_path)
+            env.load_from_pip_file(packager_path / pip_requirements_rel_path)
+        return ModelRuntime(
+            name=name, env=env, imports=list(map(pathlib.PurePosixPath, loaded_dict["imports"])), loading_from_file=True
+        )

snowflake/ml/model/custom_model.py

@@ -149,7 +149,9 @@ class CustomModel:
         context: A ModelContext object showing sub-models and artifacts related to this model.
     """
 
-    def __init__(self, context: ModelContext) -> None:
+    def __init__(self, context: Optional[ModelContext] = None) -> None:
+        if context is None:
+            context = ModelContext()
         self.context = context
         for method in self._get_infer_methods():
             _validate_predict_function(method)

snowflake/ml/model/type_hints.py

@@ -19,6 +19,8 @@ from snowflake.ml.model import deploy_platforms
 from snowflake.ml.model._signatures import core
 
 if TYPE_CHECKING:
+    import catboost
+    import lightgbm
     import mlflow
     import numpy as np
     import pandas as pd
@@ -33,7 +35,6 @@ if TYPE_CHECKING:
     import snowflake.ml.model.custom_model
     import snowflake.ml.model.models.huggingface_pipeline
     import snowflake.ml.model.models.llm
-    import snowflake.ml.model.models.sentence_transformers
     import snowflake.snowpark
     from snowflake.ml.modeling.framework import base  # noqa: F401
 
@@ -69,6 +70,9 @@ _DataType = TypeVar("_DataType", bound=SupportedDataType)
 CustomModelType = TypeVar("CustomModelType", bound="snowflake.ml.model.custom_model.CustomModel")
 
 SupportedRequireSignatureModelType = Union[
+    "catboost.CatBoost",
+    "lightgbm.LGBMModel",
+    "lightgbm.Booster",
     "snowflake.ml.model.custom_model.CustomModel",
     "sklearn.base.BaseEstimator",
     "sklearn.pipeline.Pipeline",
@@ -85,7 +89,6 @@ SupportedNoSignatureRequirementsModelType = Union[
     "transformers.Pipeline",
     "sentence_transformers.SentenceTransformer",
     "snowflake.ml.model.models.huggingface_pipeline.HuggingFacePipelineModel",
-    "snowflake.ml.model.models.sentence_transformers.SentenceTransformer",
     "snowflake.ml.model.models.llm.LLM",
 ]
 
@@ -98,11 +101,14 @@ Here is all acceptable types of Snowflake native model packaging and its handler
 
 | Type                            | Handler File | Handler             |
 |---------------------------------|--------------|---------------------|
+| catboost.CatBoost       | catboost.py   | _CatBoostModelHandler    |
 | snowflake.ml.model.custom_model.CustomModel | custom.py    | _CustomModelHandler |
 | sklearn.base.BaseEstimator      | sklearn.py   | _SKLModelHandler    |
 | sklearn.pipeline.Pipeline       | sklearn.py   | _SKLModelHandler    |
 | xgboost.XGBModel       | xgboost.py   | _XGBModelHandler    |
 | xgboost.Booster        | xgboost.py   | _XGBModelHandler    |
+| lightgbm.LGBMModel       | lightgbm.py   | _LGBMModelHandler    |
+| lightgbm.Booster        | lightgbm.py   | _LGBMModelHandler    |
 | snowflake.ml.framework.base.BaseEstimator      | snowmlmodel.py   | _SnowMLModelHandler    |
 | torch.nn.Module      | pytroch.py   | _PyTorchHandler    |
 | torch.jit.ScriptModule      | torchscript.py   | _TorchScriptHandler    |
@@ -114,8 +120,10 @@ Here is all acceptable types of Snowflake native model packaging and its handler
 """
 
 SupportedModelHandlerType = Literal[
+    "catboost",
     "custom",
     "huggingface_pipeline",
+    "lightgbm",
     "mlflow",
     "pytorch",
     "sentence_transformers",
@@ -225,6 +233,11 @@ class BaseModelSaveOption(TypedDict):
     method_options: NotRequired[Dict[str, ModelMethodSaveOptions]]
 
 
+class CatBoostModelSaveOptions(BaseModelSaveOption):
+    target_methods: NotRequired[Sequence[str]]
+    cuda_version: NotRequired[str]
+
+
 class CustomModelSaveOption(BaseModelSaveOption):
     cuda_version: NotRequired[str]
 
@@ -238,6 +251,10 @@ class XGBModelSaveOptions(BaseModelSaveOption):
     cuda_version: NotRequired[str]
 
 
+class LGBMModelSaveOptions(BaseModelSaveOption):
+    target_methods: NotRequired[Sequence[str]]
+
+
 class SNOWModelSaveOptions(BaseModelSaveOption):
     target_methods: NotRequired[Sequence[str]]
 
@@ -279,7 +296,9 @@ class LLMSaveOptions(BaseModelSaveOption):
 
 ModelSaveOption = Union[
     BaseModelSaveOption,
+    CatBoostModelSaveOptions,
     CustomModelSaveOption,
+    LGBMModelSaveOptions,
     SKLModelSaveOptions,
     XGBModelSaveOptions,
     SNOWModelSaveOptions,

snowflake/ml/modeling/_internal/estimator_utils.py

@@ -195,21 +195,26 @@ def handle_inference_result(
     shape = transformed_numpy_array.shape
     if len(shape) > 1:
         if shape[1] != len(output_cols):
-            # HeterogeneousEnsemble's transform method produce results with variying shapes
-            # from (n_samples, n_estimators) to (n_samples, n_estimators * n_classes).
-            # It is hard to predict the response shape without using fragile introspection logic.
-            # So, to avoid that we are packing the results into a dataframe of shape (n_samples, 1) with
-            # each element being a list.
-            if len(output_cols) != 1:
-                raise TypeError(
-                    "expected_output_cols must be same length as transformed array or should be of length 1."
-                    f"Currently expected_output_cols shape is {len(output_cols)}, "
-                    f"transformed array shape is {shape}. "
-                )
+            # Within UDF, it is not feasible to change the output cols because we need to
+            # query the output cols after UDF by the expected output cols
             if not within_udf:
+                # The following lines are to generate the output cols to match the length of
+                # transformed_numpy_array
                 actual_output_cols = []
                 for i in range(shape[1]):
                     actual_output_cols.append(f"{output_cols[0]}_{i}")
                 output_cols = actual_output_cols
+            else:
+                # HeterogeneousEnsemble's transform method produce results with varying shapes
+                # from (n_samples, n_estimators) to (n_samples, n_estimators * n_classes).
+                # It is hard to predict the response shape without using fragile introspection logic.
+                # So, to avoid that we are packing the results into a dataframe of shape (n_samples, 1) with
+                # each element being a list.
+                if len(output_cols) != 1:
+                    raise TypeError(
+                        "expected_output_cols must be same length as transformed array or should be of length 1."
+                        f"Currently expected_output_cols shape is {len(output_cols)}, "
+                        f"transformed array shape is {shape}. "
+                    )
 
     return transformed_numpy_array, output_cols

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

@@ -99,7 +99,10 @@ class PandasTransformHandlers:
                 original_exception=ValueError(
                     "The feature names should match with those that were passed during fit.\n"
                     f"Features seen during fit call but not present in the input: {missing_features}\n"
-                    f"Features in the input dataframe : {input_cols}\n"
+                    f"Features specified with `input_cols` in estimator "
+                    f"{self.estimator.__class__.__name__} in the input dataframe: {input_cols}\n"
+                    f"In your input dataset for current method '{inference_method}', the features are:"
+                    f" {features_in_dataset}."
                 ),
             )
         input_df = dataset[columns_to_select]

snowflake/ml/modeling/_internal/model_specifications.py

@@ -29,7 +29,7 @@ class SKLearnModelSpecifications(ModelSpecifications):
         ]
 
         # A change from previous implementation.
-        # When reusing the Sprocs for all the fit() call in the session, the static dpendencies list should include
+        # When reusing the Sprocs for all the fit() call in the session, the static dependencies list should include
         # all the possible dependencies required during the lifetime.
 
         # Include XGBoost in the dependencies if it is installed.
@@ -67,10 +67,12 @@ class XGBoostModelSpecifications(ModelSpecifications):
 class LightGBMModelSpecifications(ModelSpecifications):
     def __init__(self) -> None:
         import lightgbm
+        import sklearn
 
         imports: List[str] = ["lightgbm"]
         pkgDependencies: List[str] = [
             f"numpy=={np.__version__}",
+            f"scikit-learn=={sklearn.__version__}",
             f"lightgbm=={lightgbm.__version__}",
             f"cloudpickle=={cp.__version__}",
         ]