Finetune Controller

Finetune Controller is a robust and flexible system designed to manage and streamline the fine-tuning of machine learning models on Kubernetes, particularly within OpenShift clusters. This project leverages modern tools and workflows, enabling efficient development and deployment processes for AI-driven applications.

Features

• Local Development: Get started quickly with a streamlined setup process using uv, a high-performance Python package and project manager.
• OpenShift Integration: Simplify deployment and scaling with OpenShift-specific configurations and GPU support for intensive workloads.
• MongoDB Backend: Seamlessly connect to a local or cluster-based MongoDB database.
• Extensibility: Easily integrate with the Kubeflow Training Operator and other components for advanced workflows.

Getting Started

If the cluster is already set up continue else follow the cluster setup instructions here

Prereqs

1. Recommend using uv, an extremely fast Python package and project manager

   pip install uv

2. A container engine such as Docker or Podman

Install

1. Create virtual environment and install dependencies

   uv sync

2. Start a local developement mongo database (or connect to one on cluster with port-forward)

   Local

   docker run -d --rm --name mongodb \
       -e MONGODB_INITDB_ROOT_USERNAME="default-user" \
       -e MONGODB_INITDB_ROOT_PASSWORD="admin123456789" \
       -e MONGODB_INITDB_DATABASE="finetune" \
       -p 27017:27017 \
       mongodb/mongodb-community-server:latest

   you can port-forward this connection to your local machine

   oc port-forward service/mongodb-community-server 27017:27017 -n <namespace>

3. Connect to the Openshift cluster with the cli login command oc login. If cluster not already set up follow these steps

4. Create a project level .env file (see .env.example) and update the variables.

   cp .env.example .env

5. Make sure the virtual environment is activated and start the local finetuning controller application.

   source .venv/bin/activate
   
   uvicorn app.main:app --reload

This will:

• Start MongoDB with the required configuration
• Build and start the FastAPI server
• Make the application available at http://localhost:8000

Development and Contributing

Setup pre-commit to keep linting and code styling up to standard.

uv sync
pre-commit install

Setup OpenShift Cluster Resources

Create default project

Name can be descriptive for these examples we will use finetune-controller

oc new-project finetune-controller

Create Kubeflow project

oc new-project kubeflow

Install Kubeflow training operator

kubectl apply --server-side -k "github.com/kubeflow/training-operator.git/manifests/overlays/standalone?ref=v1.8.1"

Install Kueue

Requires Kubernetes 1.29 or newer

Follow the latest docs

Install a released version

kubectl apply --server-side -f https://github.com/kubernetes-sigs/kueue/releases/download/v0.10.1/manifests.yaml

To wait for Kueue to be fully available, run:

kubectl wait deploy/kueue-controller-manager -nkueue-system --for=condition=available --timeout=5m

Restart pods

kubectl delete pods -lcontrol-plane=controller-manager -nkueue-system

First update the namepspace for the crd LocalQueue object in default-user-queue.yaml. default namepsace: "default"

yq e '.metadata.namespace = "finetune-controller"' -i crds/kueue/default-user-queue.yaml

Apply the default CRD config for Kueue or update by following their docs

kubectl apply -f crds/kueue/

Install mongodb server

Example configuration. do properly configure for production

oc new-app -e MONGODB_INITDB_ROOT_USERNAME="default-user" -e MONGODB_INITDB_ROOT_PASSWORD="admin123456789" -e MONGODB_INITDB_DATABASE="finetune"  mongodb/mongodb-community-server:latest --namespace finetune-controller

Add GPU nodes to ROSA cluster

Go to your cluster on redhat console admin dashboard. Add a machine pool of your choosing with the following configuration:

Taints

key: nvidia.com/gpu
value: <machine pool type or other>
effect: NoSchedule

Node Labels

Key: cluster-api/accelerator
Value: <gpu type e.g. V100 or empty>

Setup AWS Secret

Example aws config

# aws_credentials.yaml
apiVersion: v1
data:
  AWS_ACCESS_KEY_ID: |base64 encoded secret
  AWS_SECRET_ACCESS_KEY: |base64 encoded secret
  AWS_REGION: |base64 encoded string
kind: Secret
metadata:
  name: aws-credentials
type: Opaque

Example for base 64 command in terminal

echo -n "VALUE" | base64

Setup Pull secrets

Example docker pull secret config

# pull_secret.yaml
apiVersion: v1
data:
  .dockerconfigjson: ...
kind: Secret
metadata:
  name: cr-pull-secret
type: kubernetes.io/dockerconfigjson

Apply these secrets

oc apply -f aws-credentials.yaml -n finetune-controller

Install Finetune Controller On OpenShift

1. Create a .env.production file and update the defaults. For this example set MONGODB_URL=mongodb://mongodb-community-server.finetune-controller.svc.cluster.local:27017

   cp .env.example .env.production

2. create the application

   oc new-app --strategy=docker --binary --name finetune-controller --env-file=".env.production" --namespace finetune-controller

3. expose services and patch tls config

   oc expose deployment/finetune-controller --port=8000
   oc expose svc/finetune-controller --port=8000
   oc patch route finetune-controller --type=merge -p '{"spec":{"tls":{"termination":"edge"}}}'

4. add cluster role binding permissions to the application

5. start a build

   oc start-build finetune-controller --from-dir=. --namespace=finetune-controller

Manually Publish Updates To Finetune Controller

Publish From current project

./scripts/publish.sh

Publish From git ~HEAD

./scripts/publish_git.sh

Adding Finetuning Models

look at setting up a finetune model