Building a Kubeflow Pipeline using Docker images from a private GitLab Registry

Question

I'm relatively new to Kubeflow and I'm trying to create a pipeline that uses Docker images stored in my private GitLab registry. I've looked through the Kubeflow documentation, but I couldn't find a straightforward way to do this.

Here's what I've tried so far:

1. I created a Docker image and pushed it to my GitLab registry. I've verified that the Docker image is present in the registry by using the GitLab UI.
2. I have set up a Kubeflow instance and have basic pipelines running without any issues using publicly available images.
3. I created a GitLab registry secret in Kubernetes

I am now stuck at the point of integrating the GitLab registry with the Kubeflow pipeline and define the pipeline.

My current pipeline file which is not working:

@component
def first_op() -> str:

@dsl.pipeline(
    name='My pipeline',
    description='My machine learning pipeline'
)
def my_pipeline():
    first_generator = first_op().set_image_pull_secrets([dsl.LocalObjectReference('gitlab-registry-secret')])
    second_generator = second_op(output_first=first_generator.output).set_image_pull_secrets([dsl.LocalObjectReference('gitlab-registry-secret')])

if __name__ == '__main__':
    kfp.compiler.Compiler().compile(my_pipeline, 'my_pipeline.yaml')

Updated code:

@component(base_image=mygitlab_registry_location)
def first_op(output_path: str):
    pass

@dsl.pipeline(
    name='My pipeline',
    description='My machine learning pipeline'
)
def my_pipeline():
    first_generator = first_op(output_path = "data/xx.txt")    
    
    # Set the image pull secret for the operation
    first_generator.apply(
        lambda container_op: container_op.container.set_image_pull_secrets([dsl.LocalObjectReference('gitlab-registry-secret')])
    )


if __name__ == '__main__':
    kfp.compiler.Compiler().compile(my_pipeline, 'my_pipeline.yaml')

Currently I am having this error:

There is no input/output parameter or artifact.