xgboost.save_model() and mlflow.xgboost.log_model() methods con python and mlflow_save_model and mlflow_log_model con R respectively. These methods also add the python_function flavor puro the MLflow Models that they produce, allowing the models esatto be interpreted as generic Python functions for inference cammino mlflow.pyfunc.load_model() . This loaded PyFunc model can only be scored with DataFrame input. You can also use the mlflow.xgboost.load_model() method preciso load MLflow Models with the xgboost model flavor per native XGBoost format.

LightGBM ( lightgbm )

The lightgbm model flavor enables logging of LightGBM models in MLflow format modo the mlflow.lightgbm.save_model() and mlflow.lightgbm.log_model() methods. These methods also add the python_function flavor esatto the MLflow Models that they produce, allowing the models onesto be interpreted as generic Python functions for inference cammino mlflow.pyfunc.load_model() . This loaded PyFunc model can only be scored with DataFrame molla. You can also use the mlflow.lightgbm.load_model() method to load MLflow Models with the lightgbm model flavor per native LightGBM format.

CatBoost ( catboost )

The catboost model flavor enables logging of CatBoost models per MLflow format inizio the mlflow.catboost.save_model() and mlflow.catboost.log_model() methods. These methods also add the python_function flavor esatto the MLflow Models that they produce, allowing the models puro be interpreted as generic Python functions for inference modo mlflow.pyfunc.load_model() . You can also use the mlflow.catboost.load_model() method onesto load MLflow Models with the catboost model flavor con native CatBoost format.

Spacy( spaCy )

The spaCy model flavor enables logging of spaCy models sopra MLflow format inizio the mlflow.spacy.save_model() and mlflow.spacy.log_model() methods. Additionally, these methods add the python_function flavor sicuro the MLflow Models that they produce, allowing the models esatto be interpreted as generic Python functions for inference inizio mlflow.pyfunc.load_model() . This loaded PyFunc model can only be scored with DataFrame incentivo. You can also use the mlflow.spacy.load_model() method to load MLflow Models with the spacy model flavor sopra native spaCy format.

Fastai( fastai )

The fastai model flavor enables logging of fastai Learner models mediante MLflow format via the mlflow.fastai.save_model() and mlflow.fastai.log_model() methods. Additionally, these methods add the python_function flavor preciso the MLflow Models that they produce, allowing the models preciso be interpreted as generic Python functions for inference coraggio mlflow.pyfunc.load_model() . This loaded PyFunc model can only be scored with DataFrame molla. You can also use the mlflow.fastai.load_model() method to load MLflow Models with the fastai model flavor per native fastai format.

Statsmodels ( statsmodels )

The statsmodels model flavor enables logging of Statsmodels models in MLflow format via the mlflow.statsmodels.save_model() and mlflow.statsmodels.log_model() methods. These methods also add the python_function flavor sicuro the MLflow Models that they produce, allowing the models preciso be interpreted as generic Python functions for inference modo mlflow.pyfunc.load_model() . This loaded PyFunc model can only be scored with DataFrame incentivo. You can also use the mlflow.statsmodels.load_model() method esatto load MLflow Models with the statsmodels model flavor in native statsmodels format.

As for now, automatic logging is restricted esatto parameters, metrics and models generated by verso call preciso fit on verso statsmodels model.

Prophet ( prophet )

The date me prophet model flavor enables logging of Prophet models con MLflow format cammino the mlflow.prophet.save_model() and mlflow.prophet.log_model() methods. These methods also add the python_function flavor onesto the MLflow Models that they produce, allowing the models onesto be interpreted as generic Python functions for inference modo mlflow.pyfunc.load_model() . This loaded PyFunc model can only be scored with DataFrame molla. You can also use the mlflow.prophet.load_model() method esatto load MLflow Models with the prophet model flavor durante native prophet format.

Model Customization

While MLflow’s built-mediante model persistence utilities are convenient for packaging models from various popular ML libraries con MLflow Model format, they do not cover every use case. For example, you may want esatto use verso model from an ML library that is not explicitly supported by MLflow’s built-durante flavors. Alternatively, you may want esatto package custom inference code and giorno puro create an MLflow Model. Fortunately, MLflow provides two solutions that can be used esatto accomplish these tasks: Custom Python Models and Custom Flavors .