aiops-mlflow-02's Introduction

AIOPs-mlflow-02

commands used -

for getting user interface

mlflow ui

for running an MLproject

To run a MLproject inside a current working directory

mlflow run .

To run a MLproject inside a current working directory with existing conda env

mlflow run . --no-conda

To run a MLproject inside a current working directory with parameters specified in MLproject

mlflow run . -P param1=0.2

To run a MLproject inside from a github repository

mlflow run https://github.com/<USERNAME>/<REPO_NAME.git -P param=5.0

To export conda environment to a file

conda env export > conda.yaml

command for prediction on serving model

for windows -

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, 0.029, 0.48, 0.98, 6.2, 29, 3.33, 1.2, 0.39, 75, 0.66]]}" http://127.0.0.1:1234/invocations

On Linux and macOS

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, 0.029, 0.48, 0.98, 6.2, 29, 3.33, 1.2, 0.39, 75, 0.66]]}' http://127.0.0.1:1234/invocations