LSTM simple example

TensorFlow
,
1

Hi, I’m trying to create a simple example to work with the lstm model
It’s not working, can you tell me what the problem is?

    const model = tf.sequential();
    const LSTM = tf.layers.lstm({units: 2, returnSequences: true});
    const input = tf.input({shape: [5, 2]});
    const output = LSTM.apply(input);

    model.add(LSTM);
    model.compile({loss: 'meanSquaredError', optimizer: 'sgd'});

    console.log(JSON.stringify(input.shape));
    console.log(JSON.stringify(output.shape));

    const xs = tf.tensor2d( [ [1,2,3,4,5], [2,3,4,5,6] ] ).print();
    const ys = tf.tensor2d( [ [2,3,4,5,6], [3,4,5,6,7] ] ).print();
  
    await model.fit(xs, ys, { epochs: 200, verbose: 0, callbacks: {
        onEpochEnd: async (epoch, logs)=>{
            console.log(logs)
        }
    }});
3

Hi @gotostereo, The input to the lstm should be 3 dimensional data (batch_size,time_step,feature) but I can see that you are passing 2D data only. please refer to this for working code example. Thank You