topic Re: Is there any file size overhead when I save models using MLflow? in Data Engineering https://community.databricks.com/t5/data-engineering/is-there-any-file-size-overhead-when-i-save-models-using-mlflow/m-p/23599#M16321 <P>There shouldn't be. Generally speaking, models will be serialized according to their 'native' format for well-known libraries like Tensorflow, xgboost, sklearn, etc. Custom model will be saved with pickle. The files exist on distributed storage as artifacts. MLflow can and does log additional metadata with the model, like its schema, sample input, environment requirements, but these are tiny additional files.</P> Thu, 17 Jun 2021 19:56:25 GMT sean_owen 2021-06-17T19:56:25Z Is there any file size overhead when I save models using MLflow? https://community.databricks.com/t5/data-engineering/is-there-any-file-size-overhead-when-i-save-models-using-mlflow/m-p/23598#M16320 Wed, 16 Jun 2021 22:28:54 GMT https://community.databricks.com/t5/data-engineering/is-there-any-file-size-overhead-when-i-save-models-using-mlflow/m-p/23598#M16320 User16826992666 2021-06-16T22:28:54Z Re: Is there any file size overhead when I save models using MLflow? https://community.databricks.com/t5/data-engineering/is-there-any-file-size-overhead-when-i-save-models-using-mlflow/m-p/23599#M16321 <P>There shouldn't be. Generally speaking, models will be serialized according to their 'native' format for well-known libraries like Tensorflow, xgboost, sklearn, etc. Custom model will be saved with pickle. The files exist on distributed storage as artifacts. MLflow can and does log additional metadata with the model, like its schema, sample input, environment requirements, but these are tiny additional files.</P> Thu, 17 Jun 2021 19:56:25 GMT https://community.databricks.com/t5/data-engineering/is-there-any-file-size-overhead-when-i-save-models-using-mlflow/m-p/23599#M16321 sean_owen 2021-06-17T19:56:25Z