Function minimize

tfjs-core/src/optimizers/optimizer.ts:59–80

* Executes `f()` and minimizes the scalar output of `f()` by computing * gradients of y with respect to the list of trainable variables provided by * `varList`. If no list is provided, it defaults to all trainable variables. * * @param f The function to execute and whose output to minimi

(f: () => Scalar, returnCost = false, varList?: Variable[])

   * @doc {heading: 'Training', subheading: 'Optimizers'}
   */
  minimize(f: () => Scalar, returnCost = false, varList?: Variable[]): Scalar
      |null {
    const {value, grads} = this.computeGradients(f, varList);

    if (varList != null) {
      const gradArray: NamedTensor[] =
          varList.map(v => ({name: v.name, tensor: grads[v.name]}));
      this.applyGradients(gradArray);
    } else {
      this.applyGradients(grads);
    }

    // Dispose gradients.
    dispose(grads);

    if (returnCost) {
      return value;
    } else {
      value.dispose();
      return null;
    }
  }

  /**
   * The number of iterations that this optimizer instance has been invoked for.