Chapter 8. Training Models with MLflow

$ pip install mlflow
$ cd ~/tmp/Machine-Learning-Engineering-with-MLflow/Chapter09/gradflow/
$ export MLFLOW_TRACKING_URI=http://localhost:5000
$ mlflow db upgrade sqlite:///mlflow_db
$ make gradflow-light

$ cd ~/tmp/Machine-Learning-Engineering-with-MLflow/Chapter08/psystock-training

$ conda env create -f conda.yaml
$ conda activate pystock-data-features

$ export MLFLOW_TRACKING_URI=http://localhost:5000

$ mlflow run .
$ mlflow ui

Creating a Docker image for your training job

$ cd ~/tmp/Machine-Learning-Engineering-with-MLflow/Chapter08/psystock-training-docker

$ docker build -t psystock_docker_training_image .

// $ export TRACKING_SERVER_URI=http://host.docker.internal:5000
$ export TRACKING_SERVER_URI=localhost:5000


// Ошибка
$ docker run -p 5000:5000 -e MLflow_TRACKING_SERVER=$TRACKING_SERVER_URI psystock_docker_training_image