Error when using API: The AutoMLJob input is invalid. Target is null or empty

Question

I'm trying to setup an Automated ML Job using the API, everything seems to be in place but I get an error "Target is null or empty", but there is no field called "Target" anywhere, so I'm very confused as to what this is referring to, I spent the last 5 hours changing things but still couldn't solve this one, I see that a new Experiment has been created in Azure so the call seems to be getting through, but no Automated ML job is created due to this error

My Json looks like this:

{     "properties": {         "componentId": null,         "computeId": null,         "description": null,         "displayName": "TestPost",         "environmentId": null,         "environmentVariables": null,         "experimentName": "myExperiment",         "identity": null,         "isArchived": false,         "jobType": "AutoML",         "outputs": null,         "services": {             "Studio": {                 "endpoint": "[edited]",                 "errorMessage": null,                 "jobServiceType": "Studio",                 "nodes": null,                 "port": null,                 "properties": null,                 "status": null,             }         },         "taskDetails": {             "cvSplitColumnNames": ["make,model","year","mileage","Price_Rs"],             "featurizationSettings": {                 "blockedTransformers": null,                 "columnNameAndTypes": null,                 "datasetLanguage": null,                 "enableDnnFeaturization": false,                 "mode": "Auto",                 "transformerParams": null             },             "limitSettings": {                 "enableEarlyTermination": true,                 "exitScore": null,                 "maxConcurrentTrials": 1,                 "maxCoresPerTrial": -1,                 "maxTrials": 1000,                 "timeout": "PT30M",                 "trialTimeout": "PT30M"             },             "logVerbosity": "NotSet",             "nCrossValidations": null,             "primaryMetric": "NormalizedRootMeanSquaredError",             "targetColumnName": "Price_Rs",             "taskType": "Regression",             "testData": {                 "description": null,                 "jobInputType": "mltable",                 "mode": "ReadOnlyMount",                 "uri": "[edited]"             },             "testDataSize": null,             "trainingData": {                 "description": null,                 "jobInputType": "mltable",                 "mode": "ReadOnlyMount",                 "uri": "[edited]"             },             "trainingSettings": {                 "allowedTrainingAlgorithms": null,                 "blockedTrainingAlgorithms": null,                 "enableDnnTraining": false,                 "enableModelExplainability": true,                 "enableOnnxCompatibleModels": false,                 "enableStackEnsemble": true,                 "enableVoteEnsemble": true,                 "ensembleModelDownloadTimeout": "PT0S",                 "stackEnsembleSettings": null,             },             "validationData": {                 "description": null,                 "jobInputType": "mltable",                 "mode": "ReadOnlyMount",                 "uri": "[edited]"             },             "validationDataSize": null,             "weightColumnName": null         }     } }

Answer 1

@Ceiling Steps I see the computeId is set to null in your case.

 "computeId": null

Based on the sample request reference for input for AutoML job you need to assign compute to run the job.

Since the compute is not provided the job is unable to run or allocate compute target and hence the error.

If this answers your query, do click Accept Answer and Yes for was this answer helpful. And, if you have any further query do let us know.