deeplearn.html

<script src="https://unpkg.com/deeplearn"></script>
<script>
var math = new deeplearn.NDArrayMathGPU();
const g = new deeplearn.Graph();

// Placeholders are input containers. This is the container for where we will
// feed an input NDArray when we execute the graph.
const inputShape = [3];
const inputTensor = g.placeholder('input', inputShape);

const labelShape = [1];
const labelTensor = g.placeholder('label', labelShape);

// Variables are containers that hold a value that can be updated from
// training.
// Here we initialize the multiplier variable randomly.
const multiplier = g.variable('multiplier', deeplearn.Array2D.randNormal([1, 3]));

// Top level graph methods take Tensors and return Tensors.
const outputTensor = g.matmul(multiplier, inputTensor);
const costTensor = g.meanSquaredCost(outputTensor, labelTensor);

const learningRate = .00001;
const batchSize = 3;

const session = new deeplearn.Session(g, math);
const optimizer = new deeplearn.SGDOptimizer(learningRate);

const inputs = [
  deeplearn.Array1D.new([1.0, 2.0, 3.0]),
  deeplearn.Array1D.new([10.0, 20.0, 30.0]),
  deeplearn.Array1D.new([100.0, 200.0, 300.0])
];

const labels = [
  deeplearn.Array1D.new([4.0]),
  deeplearn.Array1D.new([40.0]),
  deeplearn.Array1D.new([400.0])
];

// Shuffles inputs and labels and keeps them mutually in sync.
const shuffledInputProviderBuilder =
  new deeplearn.InCPUMemoryShuffledInputProviderBuilder([inputs, labels]);
const [inputProvider, labelProvider] =
  shuffledInputProviderBuilder.getInputProviders();

// Maps tensors to InputProviders.
const feedEntries = [
  {tensor: inputTensor, data: inputProvider},
  {tensor: labelTensor, data: labelProvider}
];

const NUM_BATCHES = 10;
for (let i = 0; i < NUM_BATCHES; i++) {
  // Wrap session.train in a scope so the cost gets cleaned up automatically.
  math.scope(() => {
    // Train takes a cost tensor to minimize. Trains one batch. Returns the
    // average cost as a Scalar.
    const cost = session.train(
        costTensor, feedEntries, batchSize, optimizer, deeplearn.CostReduction.MEAN);

    console.log('last average cost (' + i + '): ' + cost.get());
  });
}

// Wrap session.eval in a scope so the intermediate values get cleaned up
// automatically.
math.scope((keep, track) => {
  const testInput = track(deeplearn.Array1D.new([0.1, 0.2, 0.3]));

  // session.eval can take NDArrays as input data.
  const testFeedEntries = [
    {tensor: inputTensor, data: testInput}
  ];

  const testOutput = session.eval(outputTensor, testFeedEntries);

  console.log('---inference output---');
  console.log('shape: ' + testOutput.shape);
  console.log('value: ' + testOutput.get(0));
});

// Cleanup training data.
inputs.forEach(input => input.dispose());
labels.forEach(label => label.dispose());
</script>