Wallaroo Edge Pipeline Publish for U-Net Computer Vision for Brain Segmentation
Features:
Models:
This tutorial and the assets can be downloaded as part of the Wallaroo Tutorials repository.
Wallaroo Edge Pipeline Publish for U-Net Computer Vision for Brain Segmentation
The following example uses the U-Net for brain segmentation model trained to detect lower-grade gliomas to demonstrate how to:
- Deploy the model into a Wallaroo Ops server.
- Perform a sample inferences via the Wallaroo SDK and the API calls.
- Publish the pipeline to an OCI (Open Container Initiative) registry service.
- Deploy the published pipeline to an edge device as a Wallaroo Inference Server, and perform the same inference calls.
Upon completion, the follow up tutorial U-Net for Brain Segmentation Deployment in Wallaroo demonstrates deploying the publish in an edge/multi-cloud environment and performing sample inferences through its endpoints.
Prerequisites
- A Wallaroo Community Edition or Enterprise Edition server with Edge Deployment enabled.
- For a free license of Wallaroo Community Edition, go to https://portal.wallaroo.community/.
- A x64 edge device capable of running Docker
References
- Edge Deployment Registry Guide
- Brain MRI segmentation: The original evaluation and training images. These can be used with this inference server.
Tutorial Steps
Import Libraries
The first step is to import the libraries we’ll be using. These are included by default in the Wallaroo instance’s JupyterHub service.
Verify that the following Python libraries are installed. The wallaroo library is included with the Wallaroo Ops JupyterHub instance.
pillowtorchvisionpandaspyarrowwallarooReferences
import wallaroo
from wallaroo.pipeline import Pipeline
from wallaroo.deployment_config import DeploymentConfigBuilder
from wallaroo.framework import Framework
# used to convert the Image into a numpy array
from PIL import Image
from torchvision import transforms
import pyarrow as pa
import numpy as np
import pandas as pd
import requests
# used to display dataframe information without truncating
from IPython.display import display
pd.set_option('display.max_colwidth', None)
pd.set_option('display.max_columns', None)
Connect to the Wallaroo Instance
The first step is to connect to Wallaroo through the Wallaroo client. The Python library is included in the Wallaroo install and available through the Jupyter Hub interface provided with your Wallaroo environment.
This is accomplished using the wallaroo.Client() command, which provides a URL to grant the SDK permission to your specific Wallaroo environment. When displayed, enter the URL into a browser and confirm permissions. Store the connection into a variable that can be referenced later.
If logging into the Wallaroo instance through the internal JupyterHub service, use wl = wallaroo.Client(). For more information on Wallaroo Client settings, see the Client Connection guide.
The option request_timeout provides additional time for the Wallaroo model upload process to complete.
wl = wallaroo.Client()
Create Workspace
We will create a workspace to manage our pipeline and models. The following variables will set the name of our sample workspace then set it as the current workspace.
Workspace names must be unique. The following helper function will either create a new workspace, or retrieve an existing one with the same name. Verify that a pre-existing workspace has been shared with the targeted user.
Set the variables workspace_name to ensure a unique workspace name if required.
The workspace will then be set as the Current Workspace. Model uploads and pipeline creation through the SDK are set in the current workspace.
workspace_name = "unet-detection-run-anywhere-demonstration"
model_name = "pt-unet"
model_file_name = './models/unet.pt'
pipeline_name = "pt-unet"
workspace = wl.get_workspace(name=workspace_name, create_if_not_exist=True)
wl.set_current_workspace(workspace)
{'name': 'unet-detection-run-anywhere-demonstration', 'id': 1670, 'archived': False, 'created_by': '7d603858-88e0-472e-8f71-e41094afd7ec', 'created_at': '2025-05-16T16:06:57.419502+00:00', 'models': [], 'pipelines': []}
Upload Model
The model is uploaded as a PyTorch model. This requires the input and output schemas for the model specified in Apache Arrow Schema format.
import pyarrow as pa
input_schema = pa.schema([
pa.field('input', pa.list_(
pa.list_(
pa.list_(
pa.float32(),
list_size=256
),
list_size=256
),
list_size=3
)),
])
output_schema = pa.schema([
pa.field('output', pa.list_(
pa.list_(
pa.list_(
pa.float32(),
list_size=256
),
list_size=256
),
list_size=1
)),
])
modelpath = 'models/unet.pt'
model = wl.upload_model(model_name,
model_file_name,
framework=Framework.PYTORCH,
input_schema=input_schema,
output_schema=output_schema)
model
Waiting for model loading - this will take up to 10min.
Model is pending loading to a native runtime.............................
Model is attempting loading to a native runtime.
Successful
Ready
| Name | pt-unet |
| Version | 236fac81-1b44-47a9-8d9b-0734bad106b2 |
| File Name | unet.pt |
| SHA | 8824a893cf676a68b97e36d8352cc37a1db7816f01d6afa4467b1806c4527ffa |
| Status | ready |
| Image Path | None |
| Architecture | x86 |
| Acceleration | none |
| Updated At | 2025-16-May 16:09:31 |
| Workspace id | 1670 |
| Workspace name | unet-detection-run-anywhere-demonstration |
model.config().runtime()
'onnx'
Deploy Pipeline
We create the pipeline with the wallaroo.client.build_pipeline method, and assign our model as a model pipeline step. Once complete, we will deploy the pipeline to allocate resources from the Kuberntes cluster hosting the Wallaroo Ops to the pipeline.
pipeline = wl.build_pipeline(pipeline_name)
pipeline.add_model_step(model)
| name | pt-unet |
|---|---|
| created | 2025-05-16 16:12:20.695563+00:00 |
| last_updated | 2025-05-16 16:12:20.695563+00:00 |
| deployed | (none) |
| workspace_id | 1670 |
| workspace_name | unet-detection-run-anywhere-demonstration |
| arch | None |
| accel | None |
| tags | |
| versions | bd62f455-d809-4c9c-b881-03d36183c3cb |
| steps | |
| published | False |
Next we configure the hardware we want to use for deployment. If we plan on eventually deploying to edge, this is a good way to simulate edge hardware conditions.
The pipeline is then deployed with our deployment configuration, which allocates cluter resources to the pipeline.
deployment_config = DeploymentConfigBuilder() \
.cpus(0.25).memory('1Gi') \
.build()
pipeline.deploy(deployment_config=deployment_config)
pipeline.status()
Waiting for deployment - this will take up to 45s ............ ok
{'status': 'Running',
'details': [],
'engines': [{'ip': '10.4.2.57',
'name': 'engine-7c8787897f-b7pzm',
'status': 'Running',
'reason': None,
'details': [],
'pipeline_statuses': {'pipelines': [{'id': 'pt-unet',
'status': 'Running',
'version': '0627e43f-6c8d-4cee-bfd4-e0fb1b4d95d3'}]},
'model_statuses': {'models': [{'model_version_id': 699,
'name': 'pt-unet',
'sha': '8824a893cf676a68b97e36d8352cc37a1db7816f01d6afa4467b1806c4527ffa',
'status': 'Running',
'version': '236fac81-1b44-47a9-8d9b-0734bad106b2'}]}}],
'engine_lbs': [{'ip': '10.4.2.56',
'name': 'engine-lb-5f76cc9c94-vzv77',
'status': 'Running',
'reason': None,
'details': []}],
'sidekicks': []}
Test Inference
We will perform a test inference by converting the file TCGA_CS_4944.png into a numpy array, and setting that as a row in a DataFrame for our inference request.
input_image = Image.open("./data/TCGA_CS_4944.png")
display(input_image)
# preprocess
m, s = np.mean(input_image, axis=(0, 1)), np.std(input_image, axis=(0, 1))
preprocess = transforms.Compose([
transforms.ToTensor(),
transforms.Normalize(mean=m, std=s),
])
input_tensor = preprocess(input_image)
input_batch = input_tensor.unsqueeze(0)
nimage = input_batch.detach().numpy()
nimage.shape
(1, 3, 256, 256)
nimage = input_tensor.detach().numpy()
input_data = {
"input": [nimage]
}
dataframe = pd.DataFrame(input_data)
We can now perform an inference in two ways:
- Wallaroo SDK: The method
wallaroo.pipeline.inferaccepts a DataFrame or Apache Arrow table and returns an inference result. - Wallaroo Pipline Inference URL: Deployed pipelines provide an inference URL that accepts a DataFrame or Apache Arrow table through an API call.
For this demonstration and to save space in the notebook, only the first few elements are shown.
- References
# inference via the Wallaroo SDK
result = pipeline.infer(dataframe)
# display(result)
result["out.output"][0:5]
0 [1.475215e-05, 1.463294e-05, 1.3977289e-05, 1.3887882e-05, 1.4543533e-05, 1.4483929e-05, 1.4811754e-05, 1.4811754e-05, 1.5079975e-05, 1.4781952e-05, 1.5079975e-05, 1.5169382e-05, 1.4692545e-05, 1.4811754e-05, 1.5079975e-05, 1.5884638e-05, 1.4781952e-05, 1.6331673e-05, 1.4781952e-05, 1.5109777e-05, 1.4513731e-05, 1.4215708e-05, 1.4901161e-05, 1.6242266e-05, 1.3768673e-05, 1.5467405e-05, 1.4126301e-05, 1.4990568e-05, 1.4364719e-05, 1.4066696e-05, 1.4960766e-05, 1.6510487e-05, 1.4424324e-05, 1.6331673e-05, 1.552701e-05, 1.4960766e-05, 1.4901161e-05, 1.40964985e-05, 1.4692545e-05, 1.463294e-05, 1.6242266e-05, 1.6391277e-05, 1.6510487e-05, 1.4662743e-05, 1.4692545e-05, 1.3977289e-05, 1.4394522e-05, 1.4394522e-05, 1.5974045e-05, 1.591444e-05, 1.5825033e-05, 1.4394522e-05, 1.4424324e-05, 1.4334917e-05, 1.4305115e-05, 1.4066696e-05, 1.5348196e-05, 1.5199184e-05, 1.6063452e-05, 1.4543533e-05, 1.4483929e-05, 1.4036894e-05, 1.4334917e-05, 1.4126301e-05, 1.5377998e-05, 1.5258789e-05, 1.5467405e-05, 1.41859055e-05, 1.4543533e-05, 1.4543533e-05, 1.4513731e-05, 1.4066696e-05, 1.5169382e-05, 1.4990568e-05, 1.5884638e-05, 1.4483929e-05, 1.4543533e-05, 1.4066696e-05, 1.4305115e-05, 1.4156103e-05, 1.5437603e-05, 1.5288591e-05, 1.5437603e-05, 1.424551e-05, 1.4573336e-05, 1.4543533e-05, 1.4573336e-05, 1.4066696e-05, 1.513958e-05, 1.4990568e-05, 1.5825033e-05, 1.4483929e-05, 1.4513731e-05, 1.40964985e-05, 1.424551e-05, 1.4156103e-05, 1.54078e-05, 1.5348196e-05, 1.5467405e-05, 1.4215708e-05, ...]
Name: out.output, dtype: object
# inference via the Wallaroo Pipeline Inference URL
headers = wl.auth.auth_header()
headers['Content-Type'] = 'application/json; format=pandas-records'
deploy_url = pipeline._deployment._url()
response = requests.post(
deploy_url,
headers=headers,
data=dataframe.to_json(orient="records")
)
display(pd.DataFrame(response.json()).loc[0, 'out']['output'][0:5])
[1.475215e-05, 1.463294e-05, 1.3977289e-05, 1.3887882e-05, 1.4543533e-05]
Undeploy the Pipeline
With the inference tests complete, we can undeploy the pipeline and return the resources back to the cluster.
pipeline.undeploy()
Waiting for undeployment - this will take up to 45s .................................... ok
| name | pt-unet |
|---|---|
| created | 2025-05-16 16:12:20.695563+00:00 |
| last_updated | 2025-05-16 16:12:23.179908+00:00 |
| deployed | False |
| workspace_id | 1670 |
| workspace_name | unet-detection-run-anywhere-demonstration |
| arch | x86 |
| accel | none |
| tags | |
| versions | 0627e43f-6c8d-4cee-bfd4-e0fb1b4d95d3, bd62f455-d809-4c9c-b881-03d36183c3cb |
| steps | pt-unet |
| published | False |
Publish the Pipeline for Edge Deployment
It worked! For a demo, we’ll take working once as “tested”. So now that we’ve tested our pipeline, we are ready to publish it for edge deployment.
Publishing it means assembling all of the configuration files and model assets and pushing them to an Open Container Initiative (OCI) repository set in the Wallaroo instance as the Edge Registry service. DevOps engineers then retrieve that image and deploy it through Docker, Kubernetes, or similar deployments.
See Edge Deployment Registry Guide for details on adding an OCI Registry Service to Wallaroo as the Edge Deployment Registry.
This is done through the SDK command wallaroo.pipeline.publish(deployment_config).
Publish Example
We will now publish the pipeline to our Edge Deployment Registry with the pipeline.publish(deployment_config) command.
# edge deployment
pub = pipeline.publish()
pub
Waiting for pipeline publish... It may take up to 600 sec.
... Published.blishing.
| ID | 52 | |
| Pipeline Name | pt-unet | |
| Pipeline Version | e19be4f2-dff6-4561-9783-2b6a359a1e19 | |
| Status | Published | |
| Workspace Id | 1670 | |
| Workspace Name | unet-detection-run-anywhere-demonstration | |
| Edges | ||
| Engine URL | sample.registry.example.com/uat/engines/proxy/wallaroo/ghcr.io/wallaroolabs/fitzroy-mini:v2025.1.0-main-6139 | |
| Pipeline URL | sample.registry.example.com/uat/pipelines/pt-unet:e19be4f2-dff6-4561-9783-2b6a359a1e19 | |
| Helm Chart URL | oci://sample.registry.example.com/uat/charts/pt-unet | |
| Helm Chart Reference | sample.registry.example.com/uat/charts@sha256:c656d704477b5a784cffacd50ea665393a91a2ea9fec92a7881d2145ce579a87 | |
| Helm Chart Version | 0.0.1-e19be4f2-dff6-4561-9783-2b6a359a1e19 | |
| Engine Config | {'engine': {'resources': {'limits': {'cpu': 4.0, 'memory': '3Gi'}, 'requests': {'cpu': 4.0, 'memory': '3Gi'}, 'accel': 'none', 'arch': 'x86', 'gpu': False}}, 'engineAux': {'autoscale': {'type': 'none', 'cpu_utilization': 50.0}}} | |
| User Images | [] | |
| Created By | john.hummel@wallaroo.ai | |
| Created At | 2025-05-16 16:13:21.557585+00:00 | |
| Updated At | 2025-05-16 16:13:21.557585+00:00 | |
| Replaces | ||
| Docker Run Command |
Note: Please set the EDGE_PORT, OCI_USERNAME, and OCI_PASSWORD environment variables. | |
| Helm Install Command |
Note: Please set the HELM_INSTALL_NAME, HELM_INSTALL_NAMESPACE,
OCI_USERNAME, and OCI_PASSWORD environment variables. |
Upon completion, the follow up tutorial U-Net for Brain Segmentation Deployment in Wallaroo demonstrates deploying the publish in an edge/multi-cloud environment and performing sample inferences through its endpoints.