Troubleshooting Vertex AutoML: No Predictions Beyond Class 0 and Input Type Errors with INTEGER Data

[suzuki0430]

I have been using Vertex AutoML for multiclass classification on tabular data loaded from BigQuery, but I am encountering two inexplicable behaviors and would like to ask about them.

1. Does not predict classes other than 0
I have deployed the model and am performing real-time predictions, but it does not predict any class other than 0.
The logs also show that it predicts class 0 with values close to a score of 1.

```

[{'mostLikelyClass': '0', 'confidence': 1.0, 'allScores': [{'classLabel': '0', 'score': 1.0}, {'classLabel': '1', 'score': 2.731270721742507e-12}, {'classLabel': '10', 'score': 8.777935416569033e-15}, {'classLabel': '5', 'score': 5.986683605042941e-15}, {'classLabel': '6', 'score': 1.507040093312565e-16}, {'classLabel': '2', 'score': 1.50957856171998e-18}]}]

```

For reference, I have attached the confusion matrix after training.

2. The input type becomes STRING_VALUE
All the values in this tabular data are INTEGER, but when I input numerical values for prediction, it results in an error."

[jaia]

Hello,

Thank you for contacting the Google Cloud Community.

I have gone through your reported issue, however it seems like this is an issue observed specifically at your end. It would need more specific debugging and analysis. To ensure a faster resolution and dedicated support for your issue, I kindly request you to file a support ticket by clicking here Our support team will prioritize your request and provide you with the assistance you need.

For individual support issues, it is best to utilize the support ticketing system. We appreciate your cooperation!