Skip to content

Interactive tutorial to learn about Experiment Tracking with mlflow ๐Ÿ“ˆ

Notifications You must be signed in to change notification settings

dunnkers/experiment-tracking-with-mlflow

Folders and files

NameName
Last commit message
Last commit date

Latest commit

ย 

History

22 Commits
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 

Repository files navigation

Experiment Tracking with MLflow ๐Ÿ“ˆ

Interactive guide to experiment tracking with MLflow.

Open In Colab

Set up mlflow tracking server

  1. Create an Azure Container Registry if you don't have one yet:
az acr create --resource-group <YOUR_RESOURCE_GROUP> --name <YOUR_ACR_REGISTRY_NAME> --sku Basic --admin-enabled true
  1. Fill in the .env file in this project:
    • RESOURCE_GROUP: The name of your resource group (<YOUR_RESOURCE_GROUP> from step 1)
    • ACR_NAME: The name of your Azure Container Registry (<YOUR_ACR_REGISTRY_NAME> from step 1)
    • ACR_CONTAINER_NAME: The name of your container image (e.g. mlflow_image)
    • ACR_USERNAME: The username of your ACR
    • ACR_PASSWORD: The password of your ACR
    • AKS_CLUSTER_NAME: The name of your AKS cluster (free to choose)

Note: You can find your ACR credentials with: az acr credential show --name <YOUR_ACR_REGISTRY_NAME>

  1. Export the environment variables from the .env file:
set -a
source .env
set +a
  1. Build the container image:
docker build --platform linux/amd64 . -t ${ACR_NAME}.azurecr.io/${ACR_CONTAINER_NAME}:latest
  1. Push the container image to your ACR:
az login
az acr login --name $ACR_NAME
docker push ${ACR_NAME}.azurecr.io/${ACR_CONTAINER_NAME}:latest

Note: You can check if the image is pushed to your ACR with: az acr repository list --name $ACR_NAME --output table

  1. Create an AKS cluster:
az aks create --resource-group $RESOURCE_GROUP --name $AKS_CLUSTER_NAME --node-count 1 --generate-ssh-keys
  1. Get the credentials for your AKS cluster:
az aks get-credentials --resource-group $RESOURCE_GROUP --name $AKS_CLUSTER_NAME
  1. Create a Kubernetes secret for your ACR:
kubectl create secret docker-registry acr-secret --docker-server=$ACR_NAME.azurecr.io --docker-username=$ACR_USERNAME --docker-password=$ACR_PASSWORD
  1. Apply the deployment file to your AKS cluster:
kubectl apply -f mlflow-deployment.yaml
  1. Wait for the external IP to be assigned to the service. You can check the status with:
kubectl get service mlflow-service --watch

Once the EXTERNAL-IP is assigned, you can access your MLflow instance using that IP address on port 80 (i.e., http://< EXTERNAL-IP >/).

  1. Cleanup
az group delete --name $RESOURCE_GROUP

Or, if you want to keep your resource group, you can delete the AKS cluster and/or the ACR registry:

az aks delete --resource-group $RESOURCE_GROUP --name $AKS_CLUSTER_NAME --yes --no-wait
az acr delete --resource-group $RESOURCE_GROUP --name $ACR_NAME --yes --no-wait

About

Interactive tutorial to learn about Experiment Tracking with mlflow ๐Ÿ“ˆ

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published