ads.model.framework package

Submodules

ads.model.framework.automl_model module

class ads.model.framework.automl_model.AutoMLModel(estimator: Callable, artifact_dir: str, properties: Optional[ModelProperties] = None, auth: Optional[Dict] = None, **kwargs)

Bases: FrameworkSpecificModel

AutoMLModel class for estimators from AutoML framework.

algorithm

“ensemble”, the algorithm name of the model.

Type: str

artifact_dir

Artifact directory to store the files needed for deployment.

Type: str

auth

Default authentication is set using the ads.set_auth API. To override the default, use the ads.common.auth.api_keys or ads.common.auth.resource_principal to create an authentication signer to instantiate an IdentityClient object.

Type: Dict

ds_client

The data science client used by model deployment.

Type: DataScienceClient

estimator

A trained automl estimator/model using oracle automl.

Type: Callable

framework

“oracle_automl”, the framework name of the estimator.

Type: str

hyperparameter

The hyperparameters of the estimator.

Type: dict

metadata_custom

The model custom metadata.

Type: ModelCustomMetadata

metadata_provenance

The model provenance metadata.

Type: ModelProvenanceMetadata

metadata_taxonomy

The model taxonomy metadata.

Type: ModelTaxonomyMetadata

model_artifact

This is built by calling prepare.

Type: ModelArtifact

model_deployment

A ModelDeployment instance.

Type: ModelDeployment

model_file_name

Name of the serialized model. Default to “model.pkl”.

Type: str

model_id

The model ID.

Type: str

properties

ModelProperties object required to save and deploy model.

Type: ModelProperties

runtime_info

A RuntimeInfo instance.

Type: RuntimeInfo

schema_input

Schema describes the structure of the input data.

Type: Schema

schema_output

Schema describes the structure of the output data.

Type: Schema

serialize

Whether to serialize the model to pkl file by default. If False, you need to serialize the model manually, save it under artifact_dir and update the score.py manually.

Type: bool

version

The framework version of the model.

Type: str

delete_deployment(...): Deletes the current model deployment.

deploy(..., \*\*kwargs): Deploys a model.

from_model_artifact(uri, model_file_name, artifact_dir, ..., \*\*kwargs): Loads model from the specified folder, or zip/tar archive.

from_model_catalog(model_id, model_file_name, artifact_dir, ..., \*\*kwargs): Loads model from model catalog.

introspect(...): Runs model introspection.

predict(data, ...): Returns prediction of input data run against the model deployment endpoint.

prepare(..., \*\*kwargs): Prepare and save the score.py, serialized model and runtime.yaml file.

reload(...): Reloads the model artifact files: score.py and the runtime.yaml.

save(..., \*\*kwargs): Saves model artifacts to the model catalog.

summary_status(...): Gets a summary table of the current status.

verify(data, ...): Tests if deployment works in local environment.

Examples

>>> import tempfile
>>> import logging
>>> import warnings
>>> from ads.automl.driver import AutoML
>>> from ads.automl.provider import OracleAutoMLProvider
>>> from ads.dataset.dataset_browser import DatasetBrowser
>>> from ads.model.framework.automl_model import AutoMLModel
>>> from ads.common.model_metadata import UseCaseType
>>> ds = DatasetBrowser.sklearn().open("wine").set_target("target")
>>> train, test = ds.train_test_split(test_size=0.1, random_state = 42)

>>> ml_engine = OracleAutoMLProvider(n_jobs=-1, loglevel=logging.ERROR)
>>> oracle_automl = AutoML(train, provider=ml_engine)
>>> model, baseline = oracle_automl.train(
...                model_list=['LogisticRegression', 'DecisionTreeClassifier'],
...                random_state = 42,
...                time_budget = 500
...        )

>>> automl_model.prepare(inference_conda_env=inference_conda_env, force_overwrite=True)
>>> automl_model.verify(...)
>>> automl_model.save()
>>> model_deployment = automl_model.deploy(wait_for_completion=False)

Initiates a AutoMLModel instance.

Parameters

estimator (Callable) – Any model object generated by automl framework.
artifact_dir (str) – Directory for generate artifact.
properties ((ModelProperties, optional). Defaults to None.) – ModelProperties object required to save and deploy model.
auth ((Dict, optional). Defaults to None.) – The default authetication is set using ads.set_auth API. If you need to override the default, use the ads.common.auth.api_keys or ads.common.auth.resource_principal to create appropriate authentication signer and kwargs required to instantiate IdentityClient object.

Returns

AutoMLModel instance.

Return type

AutoMLModel

Raises

TypeError – If the input model is not an AutoML model.

serialize_model(force_overwrite: Optional[bool] = False, X_sample: Optional[Union[Dict, str, List, Tuple, ndarray, Series, DataFrame]] = None, **kwargs: Dict)

Serialize and save AutoML model using pkl.

Parameters

force_overwrite ((bool, optional). Defaults to False.) – If set as True, overwrite serialized model if exists.
X_sample (Union[Dict, str, List, np.ndarray, pd.core.series.Series, pd.core.frame.DataFrame,]. Defaults to None.) – Contains model inputs such that model(X_sample) is a valid invocation of the model. Used to generate input schema.

Returns

Nothing.

Return type

None

ads.model.framework.lightgbm_model module

class ads.model.framework.lightgbm_model.LightGBMModel(estimator: Callable, artifact_dir: str, properties: Optional[ModelProperties] = None, auth: Optional[Dict] = None, **kwargs)

Bases: FrameworkSpecificModel

LightGBMModel class for estimators from Lightgbm framework.

algorithm

The algorithm of the model.

Type: str

artifact_dir

Artifact directory to store the files needed for deployment.

Type: str

auth

Default authentication is set using the ads.set_auth API. To override the default, use the ads.common.auth.api_keys or ads.common.auth.resource_principal to create an authentication signer to instantiate an IdentityClient object.

Type: Dict

ds_client

The data science client used by model deployment.

Type: DataScienceClient

estimator

A trained lightgbm estimator/model using Lightgbm.

Type: Callable

framework

“lightgbm”, the framework name of the model.

Type: str

hyperparameter

The hyperparameters of the estimator.

Type: dict

metadata_custom

The model custom metadata.

Type: ModelCustomMetadata

metadata_provenance

The model provenance metadata.

Type: ModelProvenanceMetadata

metadata_taxonomy

The model taxonomy metadata.

Type: ModelTaxonomyMetadata

model_artifact

This is built by calling prepare.

Type: ModelArtifact

model_deployment

A ModelDeployment instance.

Type: ModelDeployment

model_file_name

Name of the serialized model.

Type: str

model_id

The model ID.

Type: str

properties

ModelProperties object required to save and deploy model.

Type: ModelProperties

runtime_info

A RuntimeInfo instance.

Type: RuntimeInfo

schema_input

Schema describes the structure of the input data.

Type: Schema

schema_output

Schema describes the structure of the output data.

Type: Schema

serialize

Whether to serialize the model to pkl file by default. If False, you need to serialize the model manually, save it under artifact_dir and update the score.py manually.

Type: bool

version

The framework version of the model.

Type: str

delete_deployment(...): Deletes the current model deployment.

deploy(..., \*\*kwargs): Deploys a model.

from_model_artifact(uri, model_file_name, artifact_dir, ..., \*\*kwargs): Loads model from the specified folder, or zip/tar archive.

from_model_catalog(model_id, model_file_name, artifact_dir, ..., \*\*kwargs): Loads model from model catalog.

introspect(...): Runs model introspection.

predict(data, ...): Returns prediction of input data run against the model deployment endpoint.

prepare(..., \*\*kwargs): Prepare and save the score.py, serialized model and runtime.yaml file.

reload(...): Reloads the model artifact files: score.py and the runtime.yaml.

save(..., \*\*kwargs): Saves model artifacts to the model catalog.

summary_status(...): Gets a summary table of the current status.

verify(data, ...): Tests if deployment works in local environment.

Examples

>>> import lightgbm as lgb
>>> import tempfile
>>> from sklearn.model_selection import train_test_split
>>> from sklearn.datasets import load_iris
>>> from ads.model.framework.lightgbm_model import LightGBMModel

>>> iris = load_iris()
>>> X, y = iris.data, iris.target

>>> X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25)
>>> train = lgb.Dataset(X_train, label=y_train)
>>> param = {
...        'objective': 'multiclass', 'num_class': 3,
...        }
>>> lightgbm_estimator = lgb.train(param, train)

>>> lightgbm_model = LightGBMModel(estimator=lightgbm_estimator,
... artifact_dir=tempfile.mkdtemp())

>>> lightgbm_model.prepare(inference_conda_env="generalml_p37_cpu_v1", force_overwrite=True)
>>> lightgbm_model.reload()
>>> lightgbm_model.verify(X_test)
>>> lightgbm_model.save()
>>> model_deployment = lightgbm_model.deploy(wait_for_completion=False)
>>> lightgbm_model.predict(X_test)

Initiates a LightGBMModel instance. This class wraps the Lightgbm model as estimator. It’s primary purpose is to hold the trained model and do serialization.

Parameters

estimator – any model object generated by Lightgbm framework
artifact_dir (str) – Directory for generate artifact.
properties ((ModelProperties, optional). Defaults to None.) – ModelProperties object required to save and deploy model.
auth ((Dict, optional). Defaults to None.) – The default authetication is set using ads.set_auth API. If you need to override the default, use the ads.common.auth.api_keys or ads.common.auth.resource_principal to create appropriate authentication signer and kwargs required to instantiate IdentityClient object.

Returns

LightGBMModel instance.

Return type

LightGBMModel

Raises

TypeError – If the input model is not a Lightgbm model or not supported for serialization.:

Examples

>>> import lightgbm as lgb
>>> import tempfile
>>> from sklearn.model_selection import train_test_split
>>> from sklearn.datasets import load_iris
>>> from ads.model.framework.lightgbm_model import LightGBMModel
>>> iris = load_iris()
>>> X, y = iris.data, iris.target
>>> X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25)
>>> train = lgb.Dataset(X_train, label=y_train)
>>> param = {
... 'objective': 'multiclass', 'num_class': 3,
... }
>>> lightgbm_estimator = lgb.train(param, train)
>>> lightgbm_model = LightGBMModel(estimator=lightgbm_estimator, artifact_dir=tempfile.mkdtemp())
>>> lightgbm_model.prepare(inference_conda_env="generalml_p37_cpu_v1")
>>> lightgbm_model.verify(X_test)
>>> lightgbm_model.save()
>>> model_deployment = lightgbm_model.deploy()
>>> lightgbm_model.predict(X_test)
>>> lightgbm_model.delete_deployment()

generate_initial_types(X_sample: Any) → List

Auto generate intial types.

Parameters: X_sample ((Any)) – Train data.
Returns: Initial types.
Return type: List

serialize_model(as_onnx: bool = False, initial_types: Optional[List[Tuple]] = None, force_overwrite: bool = False, X_sample: Optional[Union[Dict, str, List, Tuple, ndarray, Series, DataFrame]] = None, **kwargs: Dict)

Serialize and save Lightgbm model using ONNX or model specific method.

Parameters

artifact_dir (str) – Directory for generate artifact.
as_onnx ((boolean, optional). Defaults to False.) – If set as True, provide initial_types or X_sample to convert into ONNX.
initial_types ((List[Tuple], optional). Defaults to None.) – Each element is a tuple of a variable name and a type.
force_overwrite ((boolean, optional). Defaults to False.) – If set as True, overwrite serialized model if exists.
X_sample (Union[Dict, str, List, np.ndarray, pd.core.series.Series, pd.core.frame.DataFrame,]. Defaults to None.) – Contains model inputs such that model(X_sample) is a valid invocation of the model. Used to generate initial_types.

Returns

Nothing.

Return type

None

to_onnx(initial_types: List[Tuple] = None, X_sample: Optional[Union[Dict, str, List, Tuple, ndarray, Series, DataFrame]] = None, **kwargs)

Produces an equivalent ONNX model of the given Lightgbm model.

Parameters

initial_types ((List[Tuple], optional). Defaults to None.) – Each element is a tuple of a variable name and a type.
X_sample (Union[Dict, str, List, np.ndarray, pd.core.series.Series, pd.core.frame.DataFrame,]. Defaults to None.) – Contains model inputs such that model(X_sample) is a valid invocation of the model. Used to generate initial_types.

Returns

An ONNX model (type

Return type

ModelProto) which is equivalent to the input Lightgbm model.

ads.model.framework.pytorch_model module

class ads.model.framework.pytorch_model.PyTorchModel(estimator: callable, artifact_dir: str, properties: Optional[ModelProperties] = None, auth: Dict = None, **kwargs)

Bases: FrameworkSpecificModel

PyTorchModel class for estimators from Pytorch framework.

algorithm

The algorithm of the model.

Type: str

artifact_dir

Artifact directory to store the files needed for deployment.

Type: str

auth

Default authentication is set using the ads.set_auth API. To override the default, use the ads.common.auth.api_keys or ads.common.auth.resource_principal to create an authentication signer to instantiate an IdentityClient object.

Type: Dict

ds_client

The data science client used by model deployment.

Type: DataScienceClient

estimator

A trained pytorch estimator/model using Pytorch.

Type: Callable

framework

“pytorch”, the framework name of the model.

Type: str

hyperparameter

The hyperparameters of the estimator.

Type: dict

metadata_custom

The model custom metadata.

Type: ModelCustomMetadata

metadata_provenance

The model provenance metadata.

Type: ModelProvenanceMetadata

metadata_taxonomy

The model taxonomy metadata.

Type: ModelTaxonomyMetadata

model_artifact

This is built by calling prepare.

Type: ModelArtifact

model_deployment

A ModelDeployment instance.

Type: ModelDeployment

model_file_name

Name of the serialized model.

Type: str

model_id

The model ID.

Type: str

properties

ModelProperties object required to save and deploy model.

Type: ModelProperties

runtime_info

A RuntimeInfo instance.

Type: RuntimeInfo

schema_input

Schema describes the structure of the input data.

Type: Schema

schema_output

Schema describes the structure of the output data.

Type: Schema

serialize

Whether to serialize the model to pkl file by default. If False, you need to serialize the model manually, save it under artifact_dir and update the score.py manually.

Type: bool

version

The framework version of the model.

Type: str

delete_deployment(...): Deletes the current model deployment.

deploy(..., \*\*kwargs): Deploys a model.

from_model_artifact(uri, model_file_name, artifact_dir, ..., \*\*kwargs): Loads model from the specified folder, or zip/tar archive.

from_model_catalog(model_id, model_file_name, artifact_dir, ..., \*\*kwargs): Loads model from model catalog.

introspect(...): Runs model introspection.

predict(data, ...): Returns prediction of input data run against the model deployment endpoint.

prepare(..., \*\*kwargs): Prepare and save the score.py, serialized model and runtime.yaml file.

reload(...): Reloads the model artifact files: score.py and the runtime.yaml.

save(..., \*\*kwargs): Saves model artifacts to the model catalog.

summary_status(...): Gets a summary table of the current status.

verify(data, ...): Tests if deployment works in local environment.

Examples

>>> torch_model = PyTorchModel(estimator=torch_estimator,
... artifact_dir=tmp_model_dir)
>>> inference_conda_env = "generalml_p37_cpu_v1"

>>> torch_model.prepare(inference_conda_env=inference_conda_env, force_overwrite=True)
>>> torch_model.reload()
>>> torch_model.verify(...)
>>> torch_model.save()
>>> model_deployment = torch_model.deploy(wait_for_completion=False)
>>> torch_model.predict(...)

Initiates a PyTorchModel instance.

Parameters

estimator (callable) – Any model object generated by pytorch framework
artifact_dir (str) – artifact directory to store the files needed for deployment.
properties ((ModelProperties, optional). Defaults to None.) – ModelProperties object required to save and deploy model.
auth ((Dict, optional). Defaults to None.) – The default authetication is set using ads.set_auth API. If you need to override the default, use the ads.common.auth.api_keys or ads.common.auth.resource_principal to create appropriate authentication signer and kwargs required to instantiate IdentityClient object.

Returns

PyTorchModel instance.

Return type

PyTorchModel

get_data_serializer(data: Union[Dict, str, List, ndarray, Series, DataFrame, torch.Tensor], data_type: str = None)

Returns serializable input data.

Parameters

data (Union[Dict, str, list, numpy.ndarray, pd.core.series.Series,) –
pd.core.frame.DataFrame – Data expected by the model deployment predict API.
torch.Tensor] – Data expected by the model deployment predict API.
data_type (str) – Type of the data.

Returns

A class containing serialized input data and original data type information.

Return type

InputDataSerializer

Raises

TypeError – if provided data type is not supported.

serialize_model(as_onnx: bool = False, force_overwrite: bool = False, X_sample: Optional[Union[Dict, str, List, Tuple, ndarray, Series, DataFrame]] = None, use_torch_script: bool = None, **kwargs) → None

Serialize and save Pytorch model using ONNX or model specific method.

Parameters

as_onnx ((bool, optional). Defaults to False.) – If set as True, convert into ONNX model.
force_overwrite ((bool, optional). Defaults to False.) – If set as True, overwrite serialized model if exists.
X_sample (Union[list, tuple, pd.Series, np.ndarray, pd.DataFrame]. Defaults to None.) – A sample of input data that will be used to generate input schema and detect onnx_args.
use_torch_script ((bool, optional). Defaults to None (If the default value has not been changed, it will be set as False).) – If set as True, the model will be serialized as a TorchScript program. Check https://pytorch.org/tutorials/beginner/saving_loading_models.html#export-load-model-in-torchscript-format for more details. If set as False, it will only save the trained model’s learned parameters, and the score.py need to be modified to construct the model class instance first. Check https://pytorch.org/tutorials/beginner/saving_loading_models.html#save-load-state-dict-recommended for more details.
**kwargs (optional params used to serialize pytorch model to onnx,) –
following (including the) – onnx_args: (tuple or torch.Tensor), default to None Contains model inputs such that model(onnx_args) is a valid invocation of the model. Can be structured either as: 1) ONLY A TUPLE OF ARGUMENTS; 2) A TENSOR; 3) A TUPLE OF ARGUMENTS ENDING WITH A DICTIONARY OF NAMED ARGUMENTS input_names: (List[str], optional). Names to assign to the input nodes of the graph, in order. output_names: (List[str], optional). Names to assign to the output nodes of the graph, in order. dynamic_axes: (dict, optional), default to None. Specify axes of tensors as dynamic (i.e. known only at run-time).

Returns

Nothing.

Return type

None

to_onnx(path: str = None, onnx_args=None, X_sample: Optional[Union[Dict, str, List, Tuple, ndarray, Series, DataFrame]] = None, input_names: List[str] = ['input'], output_names: List[str] = ['output'], dynamic_axes=None)

Exports the given Pytorch model into ONNX format.

Parameters

path (str, default to None) – Path to save the serialized model.
onnx_args ((tuple or torch.Tensor), default to None) – Contains model inputs such that model(onnx_args) is a valid invocation of the model. Can be structured either as: 1) ONLY A TUPLE OF ARGUMENTS; 2) A TENSOR; 3) A TUPLE OF ARGUMENTS ENDING WITH A DICTIONARY OF NAMED ARGUMENTS
X_sample (Union[list, tuple, pd.Series, np.ndarray, pd.DataFrame]. Defaults to None.) – A sample of input data that will be used to generate input schema and detect onnx_args.
input_names ((List[str], optional). Defaults to ["input"].) – Names to assign to the input nodes of the graph, in order.
output_names ((List[str], optional). Defaults to ["output"].) – Names to assign to the output nodes of the graph, in order.
dynamic_axes ((dict, optional). Defaults to None.) – Specify axes of tensors as dynamic (i.e. known only at run-time).

Returns

Nothing

Return type

None

Raises

AssertionError – if onnx module is not support by the current version of torch
ValueError – if X_sample is not provided if path is not provided

ads.model.framework.sklearn_model module

class ads.model.framework.sklearn_model.SklearnModel(estimator: Callable, artifact_dir: str, properties: Optional[ModelProperties] = None, auth: Optional[Dict] = None, **kwargs)

Bases: FrameworkSpecificModel

SklearnModel class for estimators from sklearn framework.

algorithm

The algorithm of the model.

Type: str

artifact_dir

Artifact directory to store the files needed for deployment.

Type: str

auth

Default authentication is set using the ads.set_auth API. To override the default, use the ads.common.auth.api_keys or ads.common.auth.resource_principal to create an authentication signer to instantiate an IdentityClient object.

Type: Dict

ds_client

The data science client used by model deployment.

Type: DataScienceClient

estimator

A trained sklearn estimator/model using scikit-learn.

Type: Callable

framework

“scikit-learn”, the framework name of the model.

Type: str

hyperparameter

The hyperparameters of the estimator.

Type: dict

metadata_custom

The model custom metadata.

Type: ModelCustomMetadata

metadata_provenance

The model provenance metadata.

Type: ModelProvenanceMetadata

metadata_taxonomy

The model taxonomy metadata.

Type: ModelTaxonomyMetadata

model_artifact

This is built by calling prepare.

Type: ModelArtifact

model_deployment

A ModelDeployment instance.

Type: ModelDeployment

model_file_name

Name of the serialized model.

Type: str

model_id

The model ID.

Type: str

properties

ModelProperties object required to save and deploy model.

Type: ModelProperties

runtime_info

A RuntimeInfo instance.

Type: RuntimeInfo

schema_input

Schema describes the structure of the input data.

Type: Schema

schema_output

Schema describes the structure of the output data.

Type: Schema

serialize

Whether to serialize the model to pkl file by default. If False, you need to serialize the model manually, save it under artifact_dir and update the score.py manually.

Type: bool

version

The framework version of the model.

Type: str

delete_deployment(...): Deletes the current model deployment.

deploy(..., \*\*kwargs): Deploys a model.

from_model_artifact(uri, model_file_name, artifact_dir, ..., \*\*kwargs): Loads model from the specified folder, or zip/tar archive.

from_model_catalog(model_id, model_file_name, artifact_dir, ..., \*\*kwargs): Loads model from model catalog.

introspect(...): Runs model introspection.

predict(data, ...): Returns prediction of input data run against the model deployment endpoint.

prepare(..., \*\*kwargs): Prepare and save the score.py, serialized model and runtime.yaml file.

reload(...): Reloads the model artifact files: score.py and the runtime.yaml.

save(..., \*\*kwargs): Saves model artifacts to the model catalog.

summary_status(...): Gets a summary table of the current status.

verify(data, ...): Tests if deployment works in local environment.

Examples

>>> import tempfile
>>> from sklearn.model_selection import train_test_split
>>> from ads.model.framework.sklearn_model import SklearnModel
>>> from sklearn.linear_model import LogisticRegression
>>> from sklearn.datasets import load_iris

>>> iris = load_iris()
>>> X, y = iris.data, iris.target
>>> X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25)
>>> sklearn_estimator = LogisticRegression()
>>> sklearn_estimator.fit(X_train, y_train)

>>> sklearn_model = SklearnModel(estimator=sklearn_estimator,
... artifact_dir=tmp_model_dir)

>>> sklearn_model.prepare(inference_conda_env="generalml_p37_cpu_v1", force_overwrite=True)
>>> sklearn_model.reload()
>>> sklearn_model.verify(X_test)
>>> sklearn_model.save()
>>> model_deployment = sklearn_model.deploy(wait_for_completion=False)
>>> sklearn_model.predict(X_test)

Initiates a SklearnModel instance.

Parameters

estimator (Callable) – Sklearn Model
artifact_dir (str) – Directory for generate artifact.
properties ((ModelProperties, optional). Defaults to None.) – ModelProperties object required to save and deploy model.
auth ((Dict, optional). Defaults to None.) – The default authetication is set using ads.set_auth API. If you need to override the default, use the ads.common.auth.api_keys or ads.common.auth.resource_principal to create appropriate authentication signer and kwargs required to instantiate IdentityClient object.

Returns

SklearnModel instance.

Return type

SklearnModel

Examples

>>> import tempfile
>>> from sklearn.model_selection import train_test_split
>>> from ads.model.framework.sklearn_model import SklearnModel
>>> from sklearn.linear_model import LogisticRegression
>>> from sklearn.datasets import load_iris

>>> iris = load_iris()
>>> X, y = iris.data, iris.target
>>> X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25)
>>> sklearn_estimator = LogisticRegression()
>>> sklearn_estimator.fit(X_train, y_train)

>>> sklearn_model = SklearnModel(estimator=sklearn_estimator, artifact_dir=tempfile.mkdtemp())
>>> sklearn_model.prepare(inference_conda_env="dataexpl_p37_cpu_v3")
>>> sklearn_model.verify(X_test)
>>> sklearn_model.save()
>>> model_deployment = sklearn_model.deploy()
>>> sklearn_model.predict(X_test)
>>> sklearn_model.delete_deployment()

generate_initial_types(X_sample: Any) → List

Auto generate intial types.

Parameters: X_sample ((Any)) – Train data.
Returns: Initial types.
Return type: List

static is_either_numerical_or_string_dataframe(data: DataFrame) → bool: Check whether all the columns are either numerical or string for dataframe.

serialize_model(as_onnx: Optional[bool] = False, initial_types: Optional[List[Tuple]] = None, force_overwrite: Optional[bool] = False, X_sample: Optional[Union[Dict, str, List, Tuple, ndarray, Series, DataFrame]] = None, **kwargs: Dict)

Serialize and save scikit-learn model using ONNX or model specific method.

Parameters

as_onnx ((bool, optional). Defaults to False.) – If set as True, provide initial_types or X_sample to convert into ONNX.
initial_types ((List[Tuple], optional). Defaults to None.) – Each element is a tuple of a variable name and a type.
force_overwrite ((bool, optional). Defaults to False.) – If set as True, overwrite serialized model if exists.
X_sample (Union[Dict, str, List, np.ndarray, pd.core.series.Series, pd.core.frame.DataFrame,]. Defaults to None.) – Contains model inputs such that model(X_sample) is a valid invocation of the model. Used to generate initial_types.

Returns

Nothing.

Return type

None

to_onnx(initial_types: List[Tuple] = None, X_sample: Optional[Union[Dict, str, List, Tuple, ndarray, Series, DataFrame]] = None, **kwargs)

Produces an equivalent ONNX model of the given scikit-learn model.

Parameters

initial_types ((List[Tuple], optional). Defaults to None.) – Each element is a tuple of a variable name and a type.
X_sample (Union[Dict, str, List, np.ndarray, pd.core.series.Series, pd.core.frame.DataFrame,]. Defaults to None.) – Contains model inputs such that model(X_sample) is a valid invocation of the model. Used to generate initial_types.

Returns

An ONNX model (type: ModelProto) which is equivalent to the input scikit-learn model.

Return type

onnx.onnx_ml_pb2.ModelProto

ads.model.framework.tensorflow_model module

class ads.model.framework.tensorflow_model.TensorFlowModel(estimator: callable, artifact_dir: str, properties: Optional[ModelProperties] = None, auth: Dict = None, **kwargs)

Bases: FrameworkSpecificModel

TensorFlowModel class for estimators from Tensorflow framework.

algorithm

The algorithm of the model.

Type: str

artifact_dir

Directory for generate artifact.

Type: str

auth

Default authentication is set using the ads.set_auth API. To override the default, use the ads.common.auth.api_keys or ads.common.auth.resource_principal to create an authentication signer to instantiate an IdentityClient object.

Type: Dict

ds_client

The data science client used by model deployment.

Type: DataScienceClient

estimator

A trained tensorflow estimator/model using Tensorflow.

Type: Callable

framework

“tensorflow”, the framework name of the model.

Type: str

hyperparameter

The hyperparameters of the estimator.

Type: dict

metadata_custom

The model custom metadata.

Type: ModelCustomMetadata

metadata_provenance

The model provenance metadata.

Type: ModelProvenanceMetadata

metadata_taxonomy

The model taxonomy metadata.

Type: ModelTaxonomyMetadata

model_artifact

This is built by calling prepare.

Type: ModelArtifact

model_deployment

A ModelDeployment instance.

Type: ModelDeployment

model_file_name

Name of the serialized model.

Type: str

model_id

The model ID.

Type: str

properties

ModelProperties object required to save and deploy model.

Type: ModelProperties

runtime_info

A RuntimeInfo instance.

Type: RuntimeInfo

schema_input

Schema describes the structure of the input data.

Type: Schema

schema_output

Schema describes the structure of the output data.

Type: Schema

serialize

Whether to serialize the model to pkl file by default. If False, you need to serialize the model manually, save it under artifact_dir and update the score.py manually.

Type: bool

version

The framework version of the model.

Type: str

delete_deployment(...): Deletes the current model deployment.

deploy(..., \*\*kwargs): Deploys a model.

from_model_artifact(uri, model_file_name, artifact_dir, ..., \*\*kwargs): Loads model from the specified folder, or zip/tar archive.

from_model_catalog(model_id, model_file_name, artifact_dir, ..., \*\*kwargs): Loads model from model catalog.

introspect(...): Runs model introspection.

predict(data, ...): Returns prediction of input data run against the model deployment endpoint.

prepare(..., \*\*kwargs): Prepare and save the score.py, serialized model and runtime.yaml file.

reload(...): Reloads the model artifact files: score.py and the runtime.yaml.

save(..., \*\*kwargs): Saves model artifacts to the model catalog.

summary_status(...): Gets a summary table of the current status.

verify(data, ...): Tests if deployment works in local environment.

Examples

>>> from ads.model.framework.tensorflow_model import TensorFlowModel
>>> import tempfile
>>> import tensorflow as tf

>>> mnist = tf.keras.datasets.mnist
>>> (x_train, y_train), (x_test, y_test) = mnist.load_data()
>>> x_train, x_test = x_train / 255.0, x_test / 255.0

>>> tf_estimator = tf.keras.models.Sequential(
...                [
...                    tf.keras.layers.Flatten(input_shape=(28, 28)),
...                    tf.keras.layers.Dense(128, activation="relu"),
...                    tf.keras.layers.Dropout(0.2),
...                    tf.keras.layers.Dense(10),
...                ]
...            )
>>> loss_fn = tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True)
>>> tf_estimator.compile(optimizer="adam", loss=loss_fn, metrics=["accuracy"])
>>> tf_estimator.fit(x_train, y_train, epochs=1)

>>> tf_model = TensorFlowModel(estimator=tf_estimator,
... artifact_dir=tempfile.mkdtemp())
>>> inference_conda_env = "generalml_p37_cpu_v1"

>>> tf_model.prepare(inference_conda_env="generalml_p37_cpu_v1", force_overwrite=True)
>>> tf_model.verify(x_test[:1])
>>> tf_model.save()
>>> model_deployment = tf_model.deploy(wait_for_completion=False)
>>> tf_model.predict(x_test[:1])

Initiates a TensorFlowModel instance.

Parameters

estimator (callable) – Any model object generated by tensorflow framework
artifact_dir (str) – Directory for generate artifact.
properties ((ModelProperties, optional). Defaults to None.) – ModelProperties object required to save and deploy model.
auth ((Dict, optional). Defaults to None.) – The default authetication is set using ads.set_auth API. If you need to override the default, use the ads.common.auth.api_keys or ads.common.auth.resource_principal to create appropriate authentication signer and kwargs required to instantiate IdentityClient object.

Returns

TensorFlowModel instance.

Return type

TensorFlowModel

get_data_serializer(data: Union[Dict, str, List, ndarray, Series, DataFrame, tf.Tensor], data_type: str = None)

Returns serializable input data.

Parameters

data (Union[Dict, str, list, numpy.ndarray, pd.core.series.Series,) –
pd.core.frame.DataFrame – Data expected by the model deployment predict API.
tf.Tensor] – Data expected by the model deployment predict API.
data_type (str) – Type of the data.

Returns

A class containing serialized input data and original data type information.

Return type

InputDataSerializer

Raises

TypeError – if provided data type is not supported.

serialize_model(as_onnx: bool = False, X_sample: Optional[Union[Dict, str, List, Tuple, ndarray, Series, DataFrame]] = None, force_overwrite: bool = False, **kwargs) → None

Serialize and save Tensorflow model using ONNX or model specific method.

Parameters

as_onnx ((bool, optional). Defaults to False.) – If set as True, convert into ONNX model.
X_sample (Union[list, tuple, pd.Series, np.ndarray, pd.DataFrame]. Defaults to None.) – A sample of input data that will be used to generate input schema and detect input_signature.
force_overwrite ((bool, optional). Defaults to False.) – If set as True, overwrite serialized model if exists.
**kwargs (optional params used to serialize tensorflow model to onnx,) –
following (including the) – input_signature: a tuple or a list of tf.TensorSpec objects). default to None. Define the shape/dtype of the input so that model(input_signature) is a valid invocation of the model. opset_version: int. Defaults to None. Used for the ONNX model.

Returns

Nothing.

Return type

None

to_onnx(path: str = None, input_signature=None, X_sample: Optional[Union[Dict, str, List, Tuple, ndarray, Series, DataFrame]] = None, opset_version=None)

Exports the given Tensorflow model into ONNX format.

Parameters

path (str, default to None) – Path to save the serialized model.
input_signature (a tuple or a list of tf.TensorSpec objects. default to None.) – Define the shape/dtype of the input so that model(input_signature) is a valid invocation of the model.
X_sample (Union[list, tuple, pd.Series, np.ndarray, pd.DataFrame]. Defaults to None.) – A sample of input data that will be used to generate input schema and detect input_signature.
opset_version (int. Defaults to None.) – The opset to be used for the ONNX model.

Returns

Nothing

Return type

None

Raises

ValueError – if path is not provided

ads.model.framework.xgboost_model module

class ads.model.framework.xgboost_model.XGBoostModel(estimator: callable, artifact_dir: str, properties: Optional[ModelProperties] = None, auth: Dict = None, **kwargs)

Bases: FrameworkSpecificModel

XGBoostModel class for estimators from xgboost framework.

algorithm

The algorithm of the model.

Type: str

artifact_dir

Artifact directory to store the files needed for deployment.

Type: str

auth

Default authentication is set using the ads.set_auth API. To override the default, use the ads.common.auth.api_keys or ads.common.auth.resource_principal to create an authentication signer to instantiate an IdentityClient object.

Type: Dict

ds_client

The data science client used by model deployment.

Type: DataScienceClient

estimator

A trained xgboost estimator/model using Xgboost.

Type: Callable

framework

“xgboost”, the framework name of the model.

Type: str

hyperparameter

The hyperparameters of the estimator.

Type: dict

metadata_custom

The model custom metadata.

Type: ModelCustomMetadata

metadata_provenance

The model provenance metadata.

Type: ModelProvenanceMetadata

metadata_taxonomy

The model taxonomy metadata.

Type: ModelTaxonomyMetadata

model_artifact

This is built by calling prepare.

Type: ModelArtifact

model_deployment

A ModelDeployment instance.

Type: ModelDeployment

model_file_name

Name of the serialized model.

Type: str

model_id

The model ID.

Type: str

properties

ModelProperties object required to save and deploy model.

Type: ModelProperties

runtime_info

A RuntimeInfo instance.

Type: RuntimeInfo

schema_input

Schema describes the structure of the input data.

Type: Schema

schema_output

Schema describes the structure of the output data.

Type: Schema

serialize

Whether to serialize the model to pkl file by default. If False, you need to serialize the model manually, save it under artifact_dir and update the score.py manually.

Type: bool

version

The framework version of the model.

Type: str

delete_deployment(...): Deletes the current model deployment.

deploy(..., \*\*kwargs): Deploys a model.

from_model_artifact(uri, model_file_name, artifact_dir, ..., \*\*kwargs): Loads model from the specified folder, or zip/tar archive.

from_model_catalog(model_id, model_file_name, artifact_dir, ..., \*\*kwargs): Loads model from model catalog.

introspect(...): Runs model introspection.

predict(data, ...): Returns prediction of input data run against the model deployment endpoint.

prepare(..., \*\*kwargs): Prepare and save the score.py, serialized model and runtime.yaml file.

reload(...): Reloads the model artifact files: score.py and the runtime.yaml.

save(..., \*\*kwargs): Saves model artifacts to the model catalog.

summary_status(...): Gets a summary table of the current status.

verify(data, ...): Tests if deployment works in local environment.

Examples

>>> import xgboost as xgb
>>> import tempfile
>>> from sklearn.datasets import make_classification
>>> from sklearn.model_selection import train_test_split
>>> from sklearn.datasets import load_iris
>>> from ads.model.framework.xgboost_model import XGBoostModel

>>> iris = load_iris()
>>> X, y = iris.data, iris.target
>>> X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25)
>>> xgboost_estimator = xgb.XGBClassifier()
>>> xgboost_estimator.fit(X_train, y_train)

>>> xgboost_model = XGBoostModel(estimator=xgboost_estimator, artifact_dir=tmp_model_dir)
>>> xgboost_model.prepare(inference_conda_env="generalml_p37_cpu_v1", force_overwrite=True)
>>> xgboost_model.reload()
>>> xgboost_model.verify(X_test)
>>> xgboost_model.save()
>>> model_deployment = xgboost_model.deploy(wait_for_completion=False)
>>> xgboost_model.predict(X_test)

Initiates a XGBoostModel instance. This class wraps the XGBoost model as estimator. It’s primary purpose is to hold the trained model and do serialization.

Parameters

estimator – XGBoostModel
artifact_dir (str) – artifact directory to store the files needed for deployment.
properties ((ModelProperties, optional). Defaults to None.) – ModelProperties object required to save and deploy model.
auth ((Dict, optional). Defaults to None.) – The default authetication is set using ads.set_auth API. If you need to override the default, use the ads.common.auth.api_keys or ads.common.auth.resource_principal to create appropriate authentication signer and kwargs required to instantiate IdentityClient object.

Returns

XGBoostModel instance.

Return type

XGBoostModel

Examples

>>> import xgboost as xgb
>>> import tempfile
>>> from sklearn.datasets import make_classification
>>> from sklearn.model_selection import train_test_split
>>> from sklearn.datasets import load_iris
>>> from ads.model.framework.xgboost_model import XGBoostModel

>>> iris = load_iris()
>>> X, y = iris.data, iris.target

>>> X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25)
>>> train = xgb.DMatrix(X_train, y_train)
>>> test = xgb.DMatrix(X_test, y_test)
>>> xgboost_estimator = XGBClassifier()
>>> xgboost_estimator.fit(X_train, y_train)
>>> xgboost_model = XGBoostModel(estimator=xgboost_estimator, artifact_dir=tempfile.mkdtemp())
>>> xgboost_model.prepare(inference_conda_env="generalml_p37_cpu_v1")
>>> xgboost_model.verify(X_test)
>>> xgboost_model.save()
>>> model_deployment = xgboost_model.deploy()
>>> xgboost_model.predict(X_test)
>>> xgboost_model.delete_deployment()

generate_initial_types(X_sample: Any) → List

Auto generate intial types.

Parameters: X_sample ((Any)) – Train data.
Returns: Initial types.
Return type: List

serialize_model(as_onnx: bool = False, initial_types: List[Tuple] = None, force_overwrite: bool = False, X_sample: Optional[Union[Dict, str, List, Tuple, ndarray, Series, DataFrame]] = None, **kwargs)

Serialize and save Xgboost model using ONNX or model specific method.

Parameters

artifact_dir (str) – Directory for generate artifact.
as_onnx ((boolean, optional). Defaults to False.) – If set as True, provide initial_types or X_sample to convert into ONNX.
initial_types ((List[Tuple], optional). Defaults to None.) – Each element is a tuple of a variable name and a type.
force_overwrite ((boolean, optional). Defaults to False.) – If set as True, overwrite serialized model if exists.
X_sample (Union[Dict, str, List, np.ndarray, pd.core.series.Series, pd.core.frame.DataFrame,]. Defaults to None.) – Contains model inputs such that model(X_sample) is a valid invocation of the model. Used to generate initial_types.

Returns

Nothing.

Return type

None

to_onnx(initial_types: List[Tuple] = None, X_sample: Union[list, tuple, DataFrame, Series, ndarray] = None, **kwargs)

Produces an equivalent ONNX model of the given Xgboost model.

Parameters

initial_types ((List[Tuple], optional). Defaults to None.) – Each element is a tuple of a variable name and a type.
X_sample (Union[Dict, str, List, np.ndarray, pd.core.series.Series, pd.core.frame.DataFrame,]. Defaults to None.) – Contains model inputs such that model(X_sample) is a valid invocation of the model. Used to generate initial_types.

Returns

An ONNX model (type: ModelProto) which is equivalent to the input xgboost model.

Return type

onnx.onnx_ml_pb2.ModelProto

ads.model.framework package

Submodules

ads.model.framework.automl_model module

ads.model.framework.lightgbm_model module

ads.model.framework.pytorch_model module

ads.model.framework.sklearn_model module

ads.model.framework.tensorflow_model module

ads.model.framework.xgboost_model module

Module contents