The shape of dict['ToFloat'] provided in model.execute(dict) must be []

When using the latest model created by automl for tensorflowjs using https://cloud.google.com/vision/automl/object-detection/docs/edge-quickstart exactly.

Has anyone else seen this issue with exporting model from automl for tensorflowjs?

When running const predictions = await model.detect(img, options);

Getting the following error:

util.js:109 Uncaught (in promise) Error: The shape of dict['ToFloat'] provided in model.execute(dict) must be [], but was [1,1280,1920,3]
    at Ev (util.js:109)
    at graph_executor.js:543
    at Array.forEach (<anonymous>)
    at e.t.checkInputShapeAndType (graph_executor.js:534)
    at e.<anonymous> (graph_executor.js:332)
    at u (runtime.js:45)
    at Generator._invoke (runtime.js:274)
    at Generator.forEach.e.<computed> [as next] (runtime.js:97)
    at Um (runtime.js:728)
    at o (runtime.js:728)

Using

"@tensorflow/tfjs": "^2.6.0",
"@tensorflow/tfjs-automl": "^1.0.0"

Full code

<script src="/node_modules/@tensorflow/tfjs/dist/tf.min.js"></script>
<script src="/node_modules/@tensorflow/tfjs-automl/dist/tf-automl.min.js"></script>
<img id="img" src="salad.jpg">
<script>
    async function run() {
        const model = await tf.automl.loadObjectDetection('/model/model.json');
        const img = document.getElementById('img');
        const options = { score: 0.5, iou: 0.5, topk: 20 };
        console.log(model.dictionary);
        const predictions = await model.detect(img, options);
        console.log(predictions);
        // Show the resulting object on the page.
        const pre = document.createElement('pre');
        pre.textContent = JSON.stringify(predictions, null, 2);
        document.body.append(pre);
    }
    run();
</script>

[pyu10055]

@ghafran This has been fixed recently on the AutoML side, I will update this thread when the changes has been deployed. You might need to retrain the model on the autoML side.

[ghafran]

This seems to be fixed now. Its working after re-training and exporting.