Github-Commit-Transfer #2
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| name: Github-Commit-Transfer | |
| on: | |
| workflow_dispatch: | |
| jobs: | |
| # Note this workflow is just to test if i am able to fetch github commit id from one job to another | |
| build-and-push-ecr-image: | |
| name: Build and push ECR image | |
| runs-on: ubuntu-latest | |
| outputs: | |
| commit_id: ${{ steps.get_commit_id.outputs.commit_id }} | |
| steps: | |
| - name: Checkout Code | |
| uses: actions/checkout@v3 | |
| - name: Install Utilities | |
| run: | | |
| sudo apt-get update | |
| sudo apt-get install -y jq unzip | |
| - name: Configure AWS credentials | |
| uses: aws-actions/configure-aws-credentials@v1 | |
| with: | |
| aws-access-key-id: ${{ secrets.AWS_ACCESS_KEY_ID }} | |
| aws-secret-access-key: ${{ secrets.AWS_SECRET_ACCESS_KEY }} | |
| aws-region: ${{ secrets.AWS_REGION }} | |
| - name: Login to Amazon ECR | |
| id: login-ecr | |
| uses: aws-actions/amazon-ecr-login@v1 | |
| - name: Get latest commit ID | |
| id: get_commit_id | |
| run: | | |
| latest_commit=$(git rev-parse HEAD) | |
| echo "::set-output name=commit_id::$latest_commit" | |
| - name: Display the commit ID | |
| run: | | |
| echo "Latest commit ID is: ${{ steps.get_commit_id.outputs.commit_id }}" | |
| launch-runner: | |
| runs-on: ubuntu-latest | |
| needs: build-and-push-ecr-image | |
| outputs: | |
| commit_id: ${{ steps.get_commit_id.outputs.commit_id }} | |
| steps: | |
| - uses: actions/checkout@v3 | |
| - uses: iterative/setup-cml@v2 | |
| - name: Display the commit ID | |
| run: | | |
| echo "Latest commit ID is: ${{ needs.outputs.commit_id }}" | |
| - name: Get latest commit ID | |
| id: get_commit_id | |
| run: | | |
| latest_commit=$(git rev-parse HEAD) | |
| echo "::set-output name=commit_id::${{ needs.get_commit_id.outputs.commit_id }}" | |
| - name: Deploy runner on AWS EC2 | |
| env: | |
| REPO_TOKEN: ${{ secrets.PERSONAL_ACCESS_TOKEN }} | |
| AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }} | |
| AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }} | |
| run: | | |
| cml runner launch \ | |
| --cloud=aws \ | |
| --name=session-08 \ | |
| --cloud-region=ap-south-1 \ | |
| --cloud-type=g4dn.xlarge \ | |
| --cloud-hdd-size=64 \ | |
| --cloud-spot \ | |
| --single \ | |
| --labels=cml-gpu \ | |
| --idle-timeout=100 | |
| train-and-report: | |
| runs-on: [self-hosted, cml-gpu] | |
| needs: launch-runner | |
| timeout-minutes: 20 | |
| # container: | |
| # image: docker://pytorch/pytorch:2.4.0-cuda12.4-cudnn9-runtime | |
| # options: --gpus all | |
| # runs-on: ubuntu-latest | |
| steps: | |
| - name: Verify EC2 Instance | |
| run: | | |
| echo "Checking instance information..." | |
| # Check if we're on EC2 | |
| TOKEN=$(curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-metadata-token-ttl-seconds: 21600") | |
| curl -H "X-aws-ec2-metadata-token: $TOKEN" http://169.254.169.254/latest/meta-data/instance-type | |
| echo "Checking system resources..." | |
| lscpu | |
| free -h | |
| df -h | |
| nvidia-smi # This will show GPU if available | |
| echo "Checking environment..." | |
| env | grep AWS || true | |
| hostname | |
| whoami | |
| pwd | |
| # Install the AWS CLI if not already available | |
| if ! command -v aws &> /dev/null; then | |
| apt-get update | |
| apt-get install -y awscli | |
| fi | |
| # Get ECR login command and execute it | |
| $(aws ecr get-login --no-include-email --region ap-south-1) | |
| aws ecr get-login-password --region ap-south-1 | docker login --username AWS --password-stdin 306093656765.dkr.ecr.ap-south-1.amazonaws.com | |
| - name: Set up AWS CLI credentials | |
| env: | |
| AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }} | |
| AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }} | |
| AWS_DEFAULT_REGION: ap-south-1 # Change to your desired region | |
| run: | | |
| # Create the AWS config and credentials files | |
| mkdir -p ~/.aws | |
| echo "[default]" > ~/.aws/config | |
| echo "region=${AWS_DEFAULT_REGION}" >> ~/.aws/config | |
| echo "[default]" > ~/.aws/credentials | |
| echo "aws_access_key_id=${AWS_ACCESS_KEY_ID}" >> ~/.aws/credentials | |
| echo "aws_secret_access_key=${AWS_SECRET_ACCESS_KEY}" >> ~/.aws/credentials | |
| - name: Test AWS CLI | |
| run: | | |
| # Now you can run any AWS CLI command | |
| aws s3 ls # Example command to list S3 buckets | |
| - name: Install jq | |
| run: | | |
| sudo apt-get update | |
| sudo apt-get install -y jq | |
| - name: Read best checkpoint file name | |
| id: read_checkpoint | |
| run: | | |
| mkdir -p model_storage | |
| touch model_storage/checkpoint_test.txt | |
| echo "CHECKPOINT_FILE=model_storage/checkpoint_test.txt" >> $GITHUB_ENV | |
| # checkpoint_file=$(head -n 1 ./model_storage/best_model_checkpoint.txt) | |
| # echo "CHECKPOINT_FILE=$checkpoint_file" >> $GITHUB_ENV | |
| - name: Display the commit ID | |
| run: | | |
| echo "Latest commit ID is: ${{ needs.outputs.commit_id }}" | |
| - name: Upload checkpoint to S3 | |
| run: | | |
| checkpoint_path="${{ env.CHECKPOINT_FILE }}" # Use the checkpoint path from the file | |
| bucket_name="mybucket-emlo-mumbai" # Change to your S3 bucket name | |
| s3_key="session-08-checkpoint/${{ needs.outputs.commit_id }}/$(basename "$checkpoint_path")" # Define S3 key | |
| echo "Uploading $checkpoint_path to s3://$bucket_name/$s3_key" | |
| aws s3 cp "$checkpoint_path" "s3://$bucket_name/$s3_key" | |
| - name: Clean previous images and containers | |
| run: | | |
| docker system prune -f |