Azure Machine Learning - endpoint failure 502 when testing the model

Question

Very basic example from ms-learn does not work!!! So frustrating

from: https://learn.microsoft.com/en-us/learn/modules/use-automated-machine-learning/

1. Created a model
2. model deployment to ACI
3. I copied the generated py code from the Consume tab into a notebook in the same service
4. get 502

Answer 1

I've experienced the same problem and I've made the case to the Microsoft support. And I've got the correct Python code as follows.

The problem come from lack of GlobalParameters in the next sentence.
input_json = json.dumps({"Inputs":{"data": x}, "GlobalParameters": 1.0})

---

endpoint = 'YOUR_ENDPOINT' #Replace with your endpoint
key = 'YOUR_KEY' #Replace with your key

import json
import requests

An array of features based on five-day weather forecast

x = [[1,1,2022,1,0,6,0,2,0.344167,0.363625,0.805833,0.160446],
[2,1,2022,1,0,0,0,2,0.363478,0.353739,0.696087,0.248539],
[3,1,2022,1,0,1,1,1,0.196364,0.189405,0.437273,0.248309],
[4,1,2022,1,0,2,1,1,0.2,0.212122,0.590435,0.160296],
[5,1,2022,1,0,3,1,1,0.226957,0.22927,0.436957,0.1869]]

Convert the array to JSON format

input_json = json.dumps({"Inputs":{"data": x}, "GlobalParameters": 1.0})

Set the content type and authentication for the request

headers = {"Content-Type":"application/json",
"Authorization":"Bearer " + key}

Send the request

response = requests.post(endpoint, input_json, headers=headers)

If we got a valid response, display the predictions

if response.status_code == 200:
y = response.json()
print("Predictions:")
for i in range(len(x)):
print (" Day: {}. Predicted rentals: {}".format(i+1, max(0, round(y["Results"][i]))))
else:
print(response)

---

best regards,

Answer 2

Additional details:

Removing:

,
    "GlobalParameters": {
    }

from the data variable (in the Endpoints/Consume code) will result in a proper result.

Answer 3

@Maciej G Is the deployed service or endpoint in a healthy state? On the Details tab of your endpoint the Deployment state should be healthy and if you also use the test tab the response should be successful.
I have followed the tutorial and have an endpoint setup which is working as expected with the code from consume tab.

Answer 4

Thanks @[IIJ Global] Yamamoto Mitsuyoshi for the fix.

It's quite frustrating that it's not true someone taking a beginners training can follow the advice: "Don't worry too much about the details of the code." (https://learn.microsoft.com/en-us/learn/modules/use-automated-machine-learning/7-deploy-model).
The persona has to either troubleshoot while the $ counter is running or give up on the exercise.

BTW it would be great to have sandboxes for AI-900 like there are for DP-900