Correct syntax error in Multiple GPUs from Scratch

Correct a syntax error where the notebook incorrectly refers to
gluon.autograd when it should refer to mxnet.autograd.

chapter07_distributed-learning/multiple-gpus-scratch.ipynb

@@ -476,12 +476,13 @@
    },
    "outputs": [],
    "source": [
+    "from mxnet import autograd\n",
     "def train_batch(batch, params, ctx, lr):\n",
     "    # split the data batch and load them on GPUs\n",
     "    data = split_and_load(batch.data[0], ctx)\n",
     "    label = split_and_load(batch.label[0], ctx)\n",
     "    # run forward on each GPU\n",
-    "    with gluon.autograd.record():\n",
+    "    with autograd.record():\n",
     "        losses = [loss(lenet(X, W), Y) \n",
     "                  for X, Y, W in zip(data, label, params)]\n",
     "    # run backward on each gpu\n",