Method apply_gradients

tensorflow/python/keras/optimizers.py:724–725

(self, grads)

    self._track_trackable(self.iterations, name='global_step')

  def apply_gradients(self, grads):
    self.optimizer.apply_gradients(grads, global_step=self.iterations)

  def get_grads(self, loss, params):
    return self.optimizer.compute_gradients(loss, params)