MR Python client, local dev, make test-e2e, using Podman

[tarilabs]

When using Podman (desktop) and testing locally the MR Python client
seems to me in order for the locally built image to be

docker build . -f Dockerfile -t kubeflow/model-registry:latest --load

with the --load argument, otheriwise:

 => CACHED [stage-1 2/3] COPY --from=builder /workspace/model-registry .                                                                                                                                                                        0.0s
WARNING: No output specified with docker-container driver. Build result will only remain in the build cache. To push result image into registry use --push or to load image into docker use --load
+ [[ -n 1 ]]
+ CLUSTER_NAME=kind
+ echo 'Creating local Kind cluster and loading image'
Creating local Kind cluster and loading image
++ kind get clusters
+ [[ kind =~ kind ]]
+ echo 'Cluster already exists, skipping creation'
Cluster already exists, skipping creation
+ kubectl config use-context kind-kind
Switched to context "kind-kind".
+ kind load docker-image -n kind kubeflow/model-registry:latest
ERROR: image: "kubeflow/model-registry:latest" not present locally

My suggestion is to have argument passing here:

model-registry/clients/python/Makefile

Line 18 in 7bbc881

cd ../../ && IMG_VERSION=${IMG_VERSION} make image/build && LOCAL=1 ./scripts/deploy_on_kind.sh

and for the make image command to accept additional parameters, in this case --load.