RHOAIENG-4146 - Add support to import a model from the ODH Model Registry via Model Build Pipeline #246

biswassri · 2024-04-09T17:59:51Z

Adding support to use model-registry and fetch model properties from it when enabled and to continue supporting the existing pipeline without MR deployed.
JIRA: RHOAIENG-4146

Description

Added a Task that fetches properties stored in model registry when the parameter model-registry-enabled is true. So far it only fetches the model source details (git repo, revision etc.) and not the containerfile details.

The added task fetch-data runs always, when model-registry-enabled is set to true then the Task connects to MR and fetches all the model related properties from the MR specified in the model-registry-hostname. The fetch-git and fetch-s3 tasks runs based on the results of fetch-model from this task. When set to false it passes on the properties from the PipelineRun file.

Below is a sample curl command used to save the model properties in the MR.

curl -X 'POST' \
  "$MR_HOSTNAME/api/model_registry/v1alpha3/model_versions" \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
  "description": "Sample for git workflow",
  "name": "2",
  "registeredModelID": "15",
  "customProperties": {
    "fetch-model": {
      "string_value": "git",
      "metadataType": "MetadataStringValue"
    },
    "git-model-repo": {
      "string_value": "https://github.com/opendatahub-io/ai-edge.git",
      "metadataType": "MetadataStringValue"
    },
    "modelRelativePath": {
      "string_value": "pipelines/models",
      "metadataType": "MetadataStringValue"
    },
    "model-dir": {
      "string_value": "tf2model",
      "metadataType": "MetadataStringValue"
    },
    "git-model-revision": {
      "string_value": "main",
      "metadataType": "MetadataStringValue"
    }
  }
}'

How Has This Been Tested?

Deploy Model Registry in cluster using the serviceRoute set to enabled state.
Once this option is set to enabled we add the MR Route under model-registry-hostname parameter to the pipeline run file
After MR operator is deployed and populated with model parameters that are require we run the pipelineRun commands:
$oc apply -k tekton/aiedge-e2e/ $oc create -f tekton/aiedge-e2e/aiedge-e2e.tensorflow-housing.pipelinerun.yaml
The remaining workflow remains the same only the pipeline run becomes shorter

You'd expect the below green pipelineRun :

Merge criteria:

The commits are squashed in a cohesive manner and have meaningful messages.
Testing instructions have been added in the PR body (for PRs involving changes that are not immediately obvious).
The developer has manually tested the changes and verified that the changes work

openshift-ci · 2024-04-09T17:59:58Z

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please assign grdryn for approval. For more information see the Kubernetes Code Review Process.

The full list of commands accepted by this bot can be found here.

Needs approval from an approver in each of these files:

OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

jackdelahunt · 2024-04-18T10:00:45Z

So was just looking at this and wanted to get a better understanding of what is happening.

Fetching from the model registry is now the default and the results from the model registry (info about model in s3 or git) are passed to the task that will then get the model? Is that correct?

Is it expected that we only use model registry now?

pipelines/tekton/aiedge-e2e/tasks/model-registry-download.yaml

grdryn · 2024-04-24T10:44:45Z

pipelines/README.md

@@ -117,6 +117,12 @@ Create a copy of the file(s) below to include the required credentials for acces
    # Linking secret is needed only if the secret from the Robot account hasn't been already applied and linked in the previous step
    $ oc secret link pipeline credentials-image-registry
    ```
+### Setup ODH Model Registry for Storing Models


Note that the CLI docs have different instructions for setting up the model registry: https://github.com/opendatahub-io/ai-edge/blob/main/docs/cli.md#installing-model-registry

The one in this PR is the link to the MR operator repo which is changing fast.
@devguyio Maybe we would need to update the CLI with the link for the MR installations because anyone using the latest changes of MR operators would see errors?

Unless there is some custom configuration for model registry, I agree that we should defer to the instruction in the model-registry-operator repo. If we notice any issues with failed deployments then it would be best for the project to file bugs / PRs to improve their instructions then us owning a duplicate set of instructions.

So, leaving this as is. We can open a PR to the CLI readme if needed to reflect the same change.

pipelines/tekton/aiedge-e2e/aiedge-e2e.pipeline.yaml

grdryn · 2024-04-24T10:52:59Z

pipelines/tekton/aiedge-e2e/aiedge-e2e.tensorflow-housing.pipelinerun.yaml

@@ -9,11 +9,15 @@ spec:
  - name: model-name


I don't think this pipelinerun example should be changed, should it? Maybe there should be a separate pipelinerun example to demonstrate the model registry integration. WDYT?

Yep adding a third example I think would be best also 👍

I agree, its a good idea. I'll add a separate pipelinerun example.

grdryn · 2024-04-24T10:58:24Z

pipelines/tekton/aiedge-e2e/tasks/model-registry-download.yaml

+        export MODEL_ID=$(curl --silent -X 'GET' \
+        "$(params.model-registry-hostname)/api/model_registry/v1alpha3/registered_models?" -H 'accept: application/json' | \


What happens if there's authentication required on the model registry endpoint?

@grdryn I don't think MR have implemented anything with authentication. Do you see anything that I am missing? Swagger for MR
My thinking was that since the MR would be deployed on the same 'local' cluster as that of the pipeline, so it should/would have access to the endpoint.

My thinking was that since the MR would be deployed on the same 'local' cluster as that of the pipeline, so it should/would have access to the endpoint.

That will probably typically be the case, although I do wonder if there will be cases where the pipelines are run on a different cluster than the model registry. I notice that you're using a Route address to access the model registry (rather than a Service address internal to the cluster), which indicates that the model registry will at least be accessible to the outside world.

I don't think MR have implemented anything with authentication. Do you see anything that I am missing?

Yeah, I'm not sure really. Authentication could be external to the API, using something like the OpenShift OAuth proxy or something else sitting in front of the API.

I'm not sure if it's 100% necessary to be able to handle authentication for this at this point, but it's something that we should at least think about.

@LaVLaS do you have any thoughts about this?

@devguyio have you thought about authentication for the model registry in the CLI at all?

@grdryn, I reached out to the MR channel about this question in slack.
I feel the istio part of it is optional so we might be okay temporarily or I might be wrong :)

I think it's fine for this PR to leave it as-is, like you say. I think we should have a follow-up story to support authentication on it though. WDYT? If you agree, could you create that in Jira?

I agree, otherwise this PR gets too complex and a lot of changes in one. Created the Jira 6703.

pipelines/tekton/aiedge-e2e/tasks/model-registry-download.yaml

grdryn · 2024-04-24T11:04:11Z

pipelines/tekton/aiedge-e2e/tasks/model-registry-download.yaml

+          echo -n "$S3-BUCKET-NAME" | tee $(results.s3-bucket-name.path)
+        fi
+        else
+          echo -n "Fetching params from PipelineRun"


Suggested change

echo -n "Fetching params from PipelineRun"

echo -n "Using params from PipelineRun"

grdryn · 2024-04-24T11:05:21Z

pipelines/tekton/aiedge-e2e/tasks/model-registry-download.yaml

+  - description: The workspace for the downloaded model uris.
+    name: workspace
+  - description: The S3 secret.
+    name: s3-secret


Are either of these workspaces actually used here?

Sorry I missed this comment earlier. Good catch, actually this is not being used, I'm gonna test with this removed.

@grdryn It turns out we need these workspaces and the pipeline fails without them with the below error: workspace binding "s3-secret" does not match any declared workspace

I'm not suggesting that they be removed from the Pipeline altogether, just from these tasks, if they're not being used within this task. See the check-model-and-containerfile-exists.yaml task for example: the s3-secret workspace is neither used nor declared.

jackdelahunt · 2024-04-24T11:13:41Z

pipelines/tekton/aiedge-e2e/aiedge-e2e.tensorflow-housing.pipelinerun.yaml

  - name: s3-bucket-name
    value: rhoai-edge-models
+  - name: model-registry-hostname
+    value: "modelregistry-sample-http-default.apps.srbiswas-cluster2.dev.datahub.redhat.com"


If you are doing another pipeline run make sure to just change the host name here

@jackdelahunt Yeah I was planning to remove it, but since we don't have any shared cluster, should i just leave it as <insert model-registry-hostname> in the example pipelinerun?

Yeah just some example value that shows what it should be without actually working.

modelregistry-sample-http-default.apps.username-cluster.dev.datahub.redhat.com or something

grdryn · 2024-04-25T23:42:28Z

The PR check failure appears to be as a result of the following error in the PipelineRun:

[pod/aiedge-e2e-bike-rentals-auto-ml-k8s75-build-container-image-pod/step-build-and-push] [1/2] STEP 5/6: COPY $MODEL_DIR /opt/app-root/src/model/
[pod/aiedge-e2e-bike-rentals-auto-ml-k8s75-build-container-image-pod/step-build-and-push] Error: building at STEP "COPY $MODEL_DIR /opt/app-root/src/model/": checking on sources under "/workspace/source/model_dir-bike-rentals-auto-ml/bike-rentals-auto-ml": copier: stat: "/model_dir-bike-rentals-auto-ml": no such file or directory

https://gcsweb-ci.apps.ci.l2s4.p1.openshiftapps.com/gcs/test-platform-results/pr-logs/pull/opendatahub-io_ai-edge/246/pull-ci-opendatahub-io-ai-edge-main-test-ai-edge/1783098369141379072/artifacts/test-ai-edge/test-ai-edge/artifacts/aiedge-e2e-bike-rentals-auto-ml-k8s75/pipelineLogs.txt

MarianMacik

Initial review.

Also, is this somehow aligned with the CLI work? Currently we have 2 approaches to Model registry:

CLI - where we store all pipeline params in the model registry
This - where we get model location from the model registry during the run of the pipeline

@LaVLaS maybe?

pipelines/tekton/aiedge-e2e/aiedge-e2e.pipeline.yaml

grdryn · 2024-05-16T11:35:36Z

pipelines/tekton/aiedge-e2e/aiedge-e2e.pipeline.yaml

+  - name: model-registry-hostname
+    type: string
+    default: ""
+    description: The servicename route for model registry 


Suggested change

description: The servicename route for model registry

description: The domain name of the model registry

This might not be a correct suggestion. Is servicename a particular route among many routes that the model registry would have?
The intention of this suggestion is to make it less specific to the idea of a "route" in OpenShift: in future it' might be exposed using the K8s Gateway API or some other method, for example.

@grdryn My idea of calling it a servicename route was that the user would know where or which endpoint to pick-up from the Openshift UI console. In my understanding model registry probably has one route to the db, so I would agree that either is fine. WDYT?

pipelines/tekton/aiedge-e2e/aiedge-e2e.pipeline.yaml

grdryn · 2024-05-16T11:43:31Z

pipelines/README.md

@@ -117,6 +117,12 @@ Create a copy of the file(s) below to include the required credentials for acces
    # Linking secret is needed only if the secret from the Robot account hasn't been already applied and linked in the previous step
    $ oc secret link pipeline credentials-image-registry
    ```
+### Setup ODH Model Registry for Storing Models
+Setting up the Model Registry using [Model Registry Operator](https://github.com/opendatahub-io/model-registry-operator?tab=readme-ov-file#running-on-the-cluster) or the ODH Operator. We are currently using the MySQL Db of MR and needs the `serviceRoute` enabled [here](https://github.com/opendatahub-io/model-registry-operator/blob/main/config/samples/mysql/modelregistry_v1alpha1_modelregistry.yaml#L17). Once this option is set to enabled we add the MR Route under `model-registry-hostname` parameter to the pipeline run file. Following this we simply run the existing pipeline create commands to start the pipeline. 


Suggested change

Setting up the Model Registry using [Model Registry Operator](https://github.com/opendatahub-io/model-registry-operator?tab=readme-ov-file#running-on-the-cluster) or the ODH Operator. We are currently using the MySQL Db of MR and needs the `serviceRoute` enabled [here](https://github.com/opendatahub-io/model-registry-operator/blob/main/config/samples/mysql/modelregistry_v1alpha1_modelregistry.yaml#L17). Once this option is set to enabled we add the MR Route under `model-registry-hostname` parameter to the pipeline run file. Following this we simply run the existing pipeline create commands to start the pipeline.

Set up the Model Registry using [Model Registry Operator](https://github.com/opendatahub-io/model-registry-operator?tab=readme-ov-file#running-on-the-cluster) or the ODH Operator.

We are currently using the MySQL Db of MR and needs the `serviceRoute` enabled [here](https://github.com/opendatahub-io/model-registry-operator/blob/main/config/samples/mysql/modelregistry_v1alpha1_modelregistry.yaml#L17).

Once this option is set to enabled we add the MR host name under `model-registry-hostname` parameter to the pipeline run file.

Following this we simply run the existing pipeline create commands to start the pipeline.

Some minor wording suggestions here, but also I've put each sentence on a separate line, to avoid the lines being so long (it can be annoying to read really long lines in a text editor/terminal, for example). When rendered by markdown, they will all be part of the same paragraph anyway, so this is just a source change.

grdryn · 2024-05-16T11:56:04Z

pipelines/README.md

+
+![MR Schema](docs/images/edge-MR.png)
+
+The current workflow supports the above schema for storing the model parameters in the MR where the parameters from the pipeline run model is stored in the Model Version as a `customProperties` with `string_value` and `metadataType`.


Suggested change

The current workflow supports the above schema for storing the model parameters in the MR where the parameters from the pipeline run model is stored in the Model Version as a `customProperties` with `string_value` and `metadataType`.

The current workflow supports the above schema for storing the model parameters in the MR, where the parameters for the PipelineRun model are stored in the Model Version as `customProperties`.

grdryn · 2024-05-16T12:03:18Z

pipelines/tekton/aiedge-e2e/tasks/model-registry-download.yaml

+          jq -r --arg MODEL_PARAM "s3-bucket-name" '.customProperties[$MODEL_PARAM].string_value')
+          echo -n "$S3-BUCKET-NAME" | tee $(results.s3-bucket-name.path)
+        fi
+        else


Suggested change

else

else

^I think this should be at the same level of indentation as the outer if condition, rather than the inner one just above it. In other words, it's odd to see the following two words at the same level of indentation:

fi else

grdryn · 2024-05-16T12:06:21Z

pipelines/tekton/aiedge-e2e/tasks/model-registry-download.yaml

+  - description: The workspace for the downloaded model uris.
+    name: workspace
+  - description: The S3 secret.
+    name: s3-secret


I'm not suggesting that they be removed from the Pipeline altogether, just from these tasks, if they're not being used within this task. See the check-model-and-containerfile-exists.yaml task for example: the s3-secret workspace is neither used nor declared.

LaVLaS · 2024-05-21T13:44:09Z

pipelines/tekton/aiedge-e2e/aiedge-e2.model-registry.tensorflow-housing.pipelinerun.yaml

+  - name: s3-bucket-name
+    value: rhoai-edge-models
+  - name: model-registry-hostname
+    value: "modelregistry-sample-http-default.apps.srbiswas-cluster2.dev.datahub.redhat.com"


Since this is specific to the cluster where the model registry is deployed, we should change this to something that is easily identifiable to a user as what needs to be changed.

Unless we can use the internal service route for model registry based on the default instructions then the default example should just break with an easily identifiable message in the value.

If we stick to the instructions from the model-registry repo for deploying modelregistry-sample, you can use this as a default

Suggested change

value: "modelregistry-sample-http-default.apps.srbiswas-cluster2.dev.datahub.redhat.com"

value: "modelregistry-sample.odh-model-registry.svc:8080"

Yes, this is an oversight. Thanks for catching this :)

MarianMacik · 2024-05-22T11:32:36Z

pipelines/tekton/aiedge-e2e/aiedge-e2e.pipeline.yaml

+      value: $(params.model-dir)
+    - name: fetch-model
+      value: $(params.fetch-model)
+    - name: modelRelativePath


Just repeating, so it doesn't get lost. This is still not fixed:

Missing:

Suggested change

- name: modelRelativePath

value: $(params.modelRelativePath)

- name: s3-bucket-name

value: $(params.s3-bucket-name)

Also, it would be good to have the parameters in the same order both in the Pipeline here and in the Task.
Another improvement can be to have the s3-bucket-name parameter without the default value in the task. This way, the pipeline would most likely fail that the pipeline is not passing the parameter to the task. But once properly linked, it won't be failing because the pipeline parameter s3-bucket-name itself has a default value of "".

Therefore, I would:

Make the order of params here and in the task definition same.

Remove any default values for the task params, so if any of the param is not set by the pipeline, it will fail.

Add the s3-bucket-name parameter here as it is still missing.

MarianMacik · 2024-05-22T11:33:49Z

pipelines/tekton/aiedge-e2e/tasks/model-registry-download.yaml

+    type: string
+  - name: model-version
+    type: string
+  #Values to be passed by pipeline run when model-register is disabled


Suggested change

#Values to be passed by pipeline run when model-register is disabled

#Values to be passed by pipeline run when model registry is disabled

MarianMacik · 2024-05-22T11:37:52Z

pipelines/tekton/aiedge-e2e/aiedge-e2.model-registry.tensorflow-housing.pipelinerun.yaml

+kind: PipelineRun
+metadata:
+  labels:
+    tekton.dev/pipeline: aiedge-e2e


Suggested change

tekton.dev/pipeline: aiedge-e2e

tekton.dev/pipeline: aiedge-e2e

model-name: tensorflow-housing

MarianMacik · 2024-05-22T11:38:58Z

pipelines/tekton/aiedge-e2e/aiedge-e2.model-registry.tensorflow-housing.pipelinerun.yaml

+  - name: model-registry-enabled
+    value: "true"


Just an idea for better readability, try to move this model-registry param to the model-registry-hostname, so it is grouped.

MarianMacik · 2024-05-22T11:49:11Z

pipelines/tekton/aiedge-e2e/aiedge-e2.model-registry.tensorflow-housing.pipelinerun.yaml

+    tekton.dev/pipeline: aiedge-e2e
+  generateName: aiedge-e2e-
+spec:
+  params:


The list of provided param values is not complete, therefore, the pipeline run is not runnable. Parameters without default values must be provided and I have identified that these are missing:

modelRelativePath

fetch-model

LaVLaS · 2024-05-28T14:53:06Z

pipelines/tekton/aiedge-e2e/tasks/model-registry-download.yaml

+        "$(params.model-registry-hostname)/api/model_registry/v1alpha3/registered_models/$MODEL_ID/versions?" -H 'accept: application/json' | \
+        jq -r --arg MODEL_VERSION "$(params.model-version)" '.items[] | select(.name == $MODEL_VERSION).id')
+
+        echo -n "MODEL ID is $MODEL_ID"


These echo -n statements are causing the output statements to be combined on a single line

MODEL ID is 15MODEL VERSION ID is 16fetching parameters from model-registry/tekton/scripts/script-0-rq6wc: line 49: export: `S3-BUCKET-NAME=': not a valid identifier

LaVLaS · 2024-05-28T15:01:26Z

pipelines/tekton/aiedge-e2e/tasks/model-registry-download.yaml

+
+      set -Eeuo pipefail
+
+      if [ "$(params.model-registry-enabled)" == "true" ]; then


If the intent of the model-registry-download Task is to be run only when fetching from the ODH model registry then this should be a "known" requirement for successful execution. We can move this logic check to the Pipeline using the Tekton when.

Inside of the task, we should only be focused on checking that there required input parameters are valid and show a clear error when the tasks execution fails

LaVLaS · 2024-05-28T15:34:42Z

pipelines/tekton/aiedge-e2e/tasks/model-registry-download.yaml

+        echo -n "MODEL ID is $MODEL_ID"
+        echo -n "MODEL VERSION ID is $MODEL_VERSION_ID"
+
+        if [[ ! -z "$MODEL_ID" ]]; then


Assuming that an empty MODEL_ID is an unrecoverable failure, we should exit with a non-zero error code to prevent any further execution of the pipeline steps since the model information is missing

LaVLaS · 2024-05-28T18:13:49Z

pipelines/tekton/aiedge-e2e/tasks/model-registry-download.yaml

+        if [[ ! -z "$MODEL_ID" ]]; then
+          echo -n "fetching parameters from model-registry"
+
+          export FETCH_MODEL=$(curl --silent -X 'GET' \


Since we are running a GET against the same endpoint to save values from the response, can we just cache the json response after the first call and use that for all of the variable assignments?

If we expect an update to the values between each query then there should be a comment to show why so that we don't optimize the calls in a future update.

LaVLaS · 2024-05-28T20:02:38Z

pipelines/tekton/aiedge-e2e/tasks/model-registry-download.yaml

+          jq -r --arg MODEL_PARAM "modelRelativePath" '.customProperties[$MODEL_PARAM].string_value')
+          echo -n "$MODEL_RELATIVE_PATH" | tee $(results.modelRelativePath.path)
+
+          export S3-BUCKET-NAME=$(curl --silent -X 'GET' \


- is not allowed in bash shell variable names. You'll need to change this to _

Suggested change

export S3-BUCKET-NAME=$(curl --silent -X 'GET' \

export S3_BUCKET_NAME=$(curl --silent -X 'GET' \

LaVLaS · 2024-05-28T20:03:10Z

pipelines/tekton/aiedge-e2e/tasks/model-registry-download.yaml

+          export S3-BUCKET-NAME=$(curl --silent -X 'GET' \
+          "$(params.model-registry-hostname)/api/model_registry/v1alpha3/model_versions/$MODEL_VERSION_ID" -H 'accept: application/json' | \
+          jq -r --arg MODEL_PARAM "s3-bucket-name" '.customProperties[$MODEL_PARAM].string_value')
+          echo -n "$S3-BUCKET-NAME" | tee $(results.s3-bucket-name.path)


See above for variable name change

Suggested change

echo -n "$S3-BUCKET-NAME" | tee $(results.s3-bucket-name.path)

echo -n "$S3_BUCKET_NAME" | tee $(results.s3-bucket-name.path)

LaVLaS · 2024-05-30T03:54:25Z

Adding a reply for historical reference.

Also, is this somehow aligned with the CLI work? Currently we have 2 approaches to Model registry:

CLI - where we store all pipeline params in the model registry

There was a discussion about adding a CLI to support this entire workflow but the scope of the work requires collaboration with the model registry in order to differentiate from the functionality included in the tekton cli.

This - where we get model location from the model registry during the run of the pipeline

From an offline conversation with @biswassri about #246 (comment), the intent was to use the model-registry as an abstraction layer for retrieving the model storage information. Long term, I think this approach is the better solution as we want to have tasks that can be used to fetch from git or s3 directly with the option of putting some type of registry abstraction task in front that can retrieve the storage information

LaVLaS · 2024-05-30T03:55:20Z

/hold

I'm putting this PR on hold while @biswassri adds changes based on recent comments

RHOAIENG-4146

9c6695e

biswassri marked this pull request as draft April 9, 2024 18:00

openshift-ci bot added the do-not-merge/work-in-progress label Apr 9, 2024

biswassri added 2 commits April 16, 2024 09:22

Updating changes

85edc56

Update MR version

53a27b2

jackdelahunt reviewed Apr 18, 2024

View reviewed changes

pipelines/tekton/aiedge-e2e/tasks/model-registry-download.yaml Show resolved Hide resolved

biswassri added 2 commits April 23, 2024 09:42

Updating Readme

a0add7f

Adding the S3 Bucket command

846a026

biswassri removed the do-not-merge/work-in-progress label Apr 23, 2024

biswassri marked this pull request as ready for review April 23, 2024 13:59

Updating Readme

38283a6

biswassri requested review from devguyio, LaVLaS and MarianMacik April 23, 2024 18:18

grdryn reviewed Apr 24, 2024

View reviewed changes

jackdelahunt reviewed Apr 24, 2024

View reviewed changes

MarianMacik reviewed Apr 26, 2024

View reviewed changes

pipelines/tekton/aiedge-e2e/aiedge-e2e.pipeline.yaml Outdated Show resolved Hide resolved

pipelines/tekton/aiedge-e2e/aiedge-e2e.pipeline.yaml Show resolved Hide resolved

pipelines/tekton/aiedge-e2e/aiedge-e2e.pipeline.yaml Outdated Show resolved Hide resolved

Changes from PR review + rebase

208016a

grdryn reviewed May 16, 2024

View reviewed changes

LaVLaS reviewed May 21, 2024

View reviewed changes

MarianMacik reviewed May 22, 2024

View reviewed changes

LaVLaS reviewed May 28, 2024

View reviewed changes

openshift-ci bot added the do-not-merge/hold label May 30, 2024

		export MODEL_ID=$(curl --silent -X 'GET' \
		"$(params.model-registry-hostname)/api/model_registry/v1alpha3/registered_models?" -H 'accept: application/json' \| \

	echo -n "Fetching params from PipelineRun"
	echo -n "Using params from PipelineRun"

	description: The servicename route for model registry
	description: The domain name of the model registry


		![MR Schema](docs/images/edge-MR.png)

		The current workflow supports the above schema for storing the model parameters in the MR where the parameters from the pipeline run model is stored in the Model Version as a `customProperties` with `string_value` and `metadataType`.

	value: "modelregistry-sample-http-default.apps.srbiswas-cluster2.dev.datahub.redhat.com"
	value: "modelregistry-sample.odh-model-registry.svc:8080"

	#Values to be passed by pipeline run when model-register is disabled
	#Values to be passed by pipeline run when model registry is disabled

	tekton.dev/pipeline: aiedge-e2e
	tekton.dev/pipeline: aiedge-e2e
	model-name: tensorflow-housing


		set -Eeuo pipefail

		if [ "$(params.model-registry-enabled)" == "true" ]; then

	export S3-BUCKET-NAME=$(curl --silent -X 'GET' \
	export S3_BUCKET_NAME=$(curl --silent -X 'GET' \

	echo -n "$S3-BUCKET-NAME" \| tee $(results.s3-bucket-name.path)
	echo -n "$S3_BUCKET_NAME" \| tee $(results.s3-bucket-name.path)

RHOAIENG-4146 - Add support to import a model from the ODH Model Registry via Model Build Pipeline #246

Are you sure you want to change the base?

RHOAIENG-4146 - Add support to import a model from the ODH Model Registry via Model Build Pipeline #246

Conversation

biswassri commented Apr 9, 2024 • edited

Description

How Has This Been Tested?

Merge criteria:

openshift-ci bot commented Apr 9, 2024

jackdelahunt commented Apr 18, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jackdelahunt Apr 24, 2024 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

biswassri Apr 24, 2024 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

biswassri May 1, 2024 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

biswassri May 2, 2024 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

biswassri Apr 24, 2024 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

grdryn commented Apr 25, 2024

MarianMacik left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

LaVLaS May 21, 2024 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

LaVLaS May 28, 2024 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

LaVLaS commented May 30, 2024

LaVLaS commented May 30, 2024

biswassri commented Apr 9, 2024 •

edited

jackdelahunt Apr 24, 2024 •

edited

biswassri Apr 24, 2024 •

edited

biswassri May 1, 2024 •

edited

biswassri May 2, 2024 •

edited

biswassri Apr 24, 2024 •

edited

LaVLaS May 21, 2024 •

edited

LaVLaS May 28, 2024 •

edited