Shulou(Shulou.com)06/01 Report--

This article mainly introduces the relevant knowledge of how to use mlflow's model registry, the content is detailed and easy to understand, the operation is simple and fast, and it has a certain reference value. I believe you will gain something after reading this article on how to use mlflow's model registry. Let's take a look at it.

What is it

This article is based on mlflow version 1.11.0

As the official website says:

The MLflow Model Registry component is a centralized model store, set of APIs, and UI, to collaboratively manage the full lifecycle of an MLflow Model. It provides model lineage (which MLflow experiment and run produced the model), model versioning, stage transitions (for example from staging to production), and annotations

Model registry is a centralized model store, apis,UI, used for full-cycle management model, it can provide a model consanguinity, model version, and model phase switching.

If model registry is not used, we have to start the service as follows:

Export MLFLOW_TRACKING_URI= http://localhost:5002 mlflow models serve-m runs:/e69aed0b22fb45debd115dfc09dbc75a/model-p 1234-- no-conda

Here we have to provide RUN_ID, that is, e69aed0b22fb45debd115dfc09dbc75a

If we use model registry, we can start the service as follows:

Export MLFLOW_TRACKING_URI= http://localhost:5002mlflow models serve-m "models:/sk-learn-random-forest-reg-model/Production"

Where sk-learn-random-forest-reg-model is the name of model, and Production is the stage phase, which will be described in more detail.

What's the specific use of that: model registry allows us to start the service without specifying RUN_ID, so that we don't need to find RUN_ID every time we start the service, so we don't need to modify the file every time we restart, which is much more convenient for algorithm personnel, and model registry divides stage logically, and can switch between stage. In this way, when we manage model, we can intuitively know which model the current algorithm service is based on.

How to operate

Let's now demonstrate how to operate model registry, assuming that we have performed several python wine.py operations according to mlflow series 1, so that we can see multiple versions of an experiment on the interface, as follows:

Click Registry model as follows:

In this way, we can register the model, and the name of the model (assuming we are wine here) can be entered by ourselves.

Click wine

Click Version3

In this way, we can switch between stage. The default stage is None, Staging indicates that it is in the preparation stage, Production indicates the online environment stage, and Archived indicates the archiving stage, that is, it is in the abandoned state.

We can also delete model: of course, you can delete a version of model specifically, as long as you click on the version, and then delete it in the same location as above.

In this way, we can start and access the service as follows. We don't need to care about the specific RUN_ID, we just need to switch from stage to Production on the interface.

# start the service mlflow models serve-m "models:/wine/Production"-p 12346-h 0.0.0.0-- no-conda# access service curl-X POST-H "Content-Type:application/json Format=pandas-split "- data'{" columns ": [" alcohol "," chlorides "," citric acid "," density "," fixed acidity "," free sulfur dioxide "," pH "," residual sugar "," sulphates "," total sulfur dioxide "," volatile acidity "]," data ": [[12.8,0.48,0.48,6.2,29,3.33,1.2,0.39,75" This is the end of http://localhost:12346/invocations 's article on "how to use mlflow's model registry". Thank you for reading! I believe you all have a certain understanding of the knowledge of "how to use mlflow's model registry". If you want to learn more, you are welcome to follow the industry information channel.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.