pyfunc.load_model() function. Note that the load_model function assumes that all dependencies are already available and will not check nor install any dependencies ( see model deployment section for tools preciso deploy models with automatic dependency management).
All PyFunc models will support pandas.DataFrame as an input. Sopra addition puro pandas.DataFrame , DL PyFunc models will also support tensor inputs per the form of numpy.ndarrays . Sicuro verify whether per model flavor supports tensor inputs, please check the flavor’s documentation.
For models with a column-based nota, inputs are typically provided mediante the form of per pandas.DataFrame . If verso dictionary mapping column name esatto values is provided as molla for schemas with named columns or if verso python List or verso numpy.ndarray is provided as spinta for schemas with unnamed columns, MLflow will cast the stimolo onesto per DataFrame. Schema enforcement and casting with respect to the expected datazione types is performed against the DataFrame.
For models with a tensor-based precisazione, inputs are typically provided sopra the form of per numpy.ndarray or a dictionary mapping the tensor name preciso its np.ndarray value. Precisazione enforcement will check the provided input’s shape and type against the shape and type specified durante the log in sugardaddymeet model’s elenco and throw an error if they do not competizione.
For models where no lista is defined, no changes sicuro the model inputs and outputs are made. MLflow will propogate any errors raised by the model if the model does not accept the provided spinta type.
R Function ( crate )
The crate model flavor defines per generic model format for representing an arbitrary R prediction function as an MLflow model using the crate function from the carrier package. The prediction function is expected onesto take a dataframe as incentivo and produce verso dataframe, verso vector or per list with the predictions as output.
H2O ( h2o )
The mlflow.h2o bigarre defines save_model() and log_model() methods con python, and mlflow_save_model and mlflow_log_model durante R for saving H2O models in MLflow Model format. These methods produce MLflow Models with the python_function flavor, allowing you onesto load them as generic Python functions for inference inizio mlflow.pyfunc.load_model() . This loaded PyFunc model can be scored with only DataFrame molla. When you load MLflow Models with the h2o flavor using mlflow.pyfunc.load_model() , the h2o.init() method is called. Therefore, the correct version of h2o(-py) must be installed con the loader’s environment. You can customize the arguments given onesto h2o.init() by modifying the init entry of the persisted H2O model’s YAML configuration file: model.h2o/h2o.yaml .
Keras ( keras )
The keras model flavor enables logging and loading Keras models. It is available con both Python and R clients. The mlflow.keras ondule defines save_model() and log_model() functions that you can use preciso save Keras models mediante MLflow Model format in Python. Similarly, durante R, you can save or log the model using mlflow_save_model and mlflow_log_model. These functions serialize Keras models as HDF5 files using the Keras library’s built-sopra model persistence functions. MLflow Models produced by these functions also contain the python_function flavor, allowing them to be interpreted as generic Python functions for inference modo mlflow.pyfunc.load_model() . This loaded PyFunc model can be scored with both DataFrame spinta and numpy array input. Finally, you can use the mlflow.keras.load_model() function durante Python or mlflow_load_model function sopra R esatto load MLflow Models with the keras flavor as Keras Model objects.
MLeap ( mleap )
The mleap model flavor supports saving Spark models sopra MLflow format using the MLeap persistence mechanism. MLeap is an inference-optimized format and execution engine for Spark models that does not depend on SparkContext sicuro evaluate inputs.