Does this really have auto diff?

I saw that for

//============================================
// Loss functions
case class SoftmaxLoss(actual: Variable, target: Variable) extends Function {
  val x: Tensor = actual.data
  val y: Tensor = target.data.T

  val shiftedLogits: Tensor = x - ns.max(x, axis = 1)
  val z: Tensor = ns.sum(ns.exp(shiftedLogits), axis = 1)
  val logProbs: Tensor = shiftedLogits - ns.log(z)
  val n: Int = x.shape.head
  val loss: Double = -ns.sum(logProbs(ns.arange(n), y)) / n

  override def forward(): Variable = Variable(Tensor(loss), Some(this))

  override def backward(gradOutput: Variable /* not used */ ): Unit = {
    val dx = ns.exp(logProbs)
    dx(ns.arange(n), y) -= 1
    dx /= n

    actual.backward(Variable(dx))
  }
}

the backward is manually implemented,which should have been automatically derived

[koen-dejonghe]

Implicitly, I think so, yes. The README has a detailed description of how auto diff works. It’s all about Variables and Functions.

[doofin]

thanks,it works! Is auto diff in this project implemented with tape based mechanism?