Question

我正在使用Tensorflow识别图像。为此，我将图像转换为张量并尝试对其进行训练，并希望保存该模型并根据一个输入图像进行预测。

下面是我使用的代码，但由于此错误而卡住了，这可能是在培训时出现的。

<apex:page id="PageId" showheader="false">
<head>
    <title>Image Classifier with TensorFlowJS</title> 
</head>
<body>
    <script src="https://cdn.jsdelivr.net/npm/@tensorflow/tfjs@0.15.1">
    </script>
</body>
<img id="imgshow" src="{!$Resource.cat}" crossorigin="anonymous" width="224" height="224" />
<button onClick="learnlinear()" lable = "Predict">Predict</button>
<script>
async function learnlinear()
{
    //img data set
    const imageHTML = document.getElementById('imgshow'); 
    console.log('imageHTML::'+imageHTML.src);

    //convert to tensor 
    const tensorImg = tf.fromPixels(imageHTML);
    tensorImg.data().then(async function (stuffTensImg){
        console.log('stuffTensImg::'+stuffTensImg.toString());
    });

    const resize_image = tf.reshape(tensorImg, [1, 224, 224, 3],'resize');
    console.log('resize_image'+resize_image);

    const model = await tf.loadModel('https://storage.googleapis.com/tfjs-models/tfjs/mobilenet_v1_0.25_224/model.json');
    console.log(':::::::'+JSON.stringify(model));

    // Use ADAM optimizer with learning rate of 0.0005 and MSE loss
    model.compile({
        optimizer: tf.train.adam(0.0005),
        loss: 'meanSquaredError',
    });
    console.log('model.compile',model.compile);

    await model.fit(resize_image, {epochs: 500});
    console.log('model.fit',model.fit);

    const saveResult = await model.save();
    console.log('saveResult',saveResult);

    model.predict(resize_image.expandDims(0)).print()
    console.log('model.predict',model.predict);
}
</script>

Error

请帮助我解决此问题或如何获取此信息。

错误：没有为“ reshape_2”提供数据。需要每个按键的数据：reshape_2

0 个答案: