diff --git a/GNNs/PyG/gcn_link_prediction.ipynb b/GNNs/PyG/gcn_link_prediction.ipynb
index 8efda3f..3d6617a 100644
--- a/GNNs/PyG/gcn_link_prediction.ipynb
+++ b/GNNs/PyG/gcn_link_prediction.ipynb
@@ -672,7 +672,7 @@
     "train_edge_neighbor_loader = conn.gds.edgeNeighborLoader(\n",
     "    v_in_feats=[\"x\"],\n",
     "    v_out_labels=[\"y\"],\n",
-    "    num_batches=5,\n",
+    "    batch_size=1000,\n",
     "    e_extra_feats=[\"is_train\",\"is_val\"],\n",
     "    output_format=\"PyG\",\n",
     "    num_neighbors=10,\n",
@@ -692,7 +692,7 @@
     "val_edge_neighbor_loader = conn.gds.edgeNeighborLoader(\n",
     "    v_in_feats=[\"x\"],\n",
     "    v_out_labels=[\"y\"],\n",
-    "    num_batches=5,\n",
+    "    batch_size=500,\n",
     "    e_extra_feats=[\"is_train\",\"is_val\"],\n",
     "    output_format=\"PyG\",\n",
     "    num_neighbors=10,\n",
diff --git a/applications/nodepiece/nodepiece.ipynb b/applications/nodepiece/nodepiece.ipynb
index 13f58b5..bac92e8 100644
--- a/applications/nodepiece/nodepiece.ipynb
+++ b/applications/nodepiece/nodepiece.ipynb
@@ -368,12 +368,12 @@
    "source": [
     "import time\n",
     "import numpy as np\n",
-    "from pyTigerGraph.gds.metrics import Accuracy\n",
+    "from pyTigerGraph.gds.metrics import Accuracy, Accumulator\n",
     "\n",
     "\n",
     "for i in range(10):\n",
     "    acc = Accuracy()\n",
-    "    epoch_loss = 0\n",
+    "    epoch_loss = Accumulator()\n",
     "    start = time.time()\n",
     "    for batch in np_loader:\n",
     "        labels = batch[\"y\"]\n",
@@ -383,21 +383,21 @@
     "        optimizer.zero_grad()\n",
     "        loss_val.backward()\n",
     "        optimizer.step()\n",
-    "        epoch_loss += loss_val.item()\n",
+    "        epoch_loss.update(loss_val.item())\n",
     "    end = time.time()\n",
     "    val_acc = Accuracy()\n",
-    "    val_epoch_loss = 0\n",
+    "    val_epoch_loss = Accumulator()\n",
     "    for val_batch in valid_loader:\n",
     "        with torch.no_grad():\n",
     "            labels = val_batch[\"y\"]\n",
     "            out = model(val_batch)\n",
     "            loss_val = loss(out, labels)\n",
     "            val_acc.update(out.argmax(dim=1), labels)\n",
-    "            val_epoch_loss += loss_val.item()\n",
-    "    print(\"EPOCH {}: {}\".format(i, epoch_loss/np_loader.num_batches), \n",
+    "            val_epoch_loss.update(loss_val.item())\n",
+    "    print(\"EPOCH {}: {}\".format(i, epoch_loss.mean), \n",
     "          \"Train Accuracy:\", acc.value, \n",
     "          \"Time:\", end-start, \n",
-    "          \"Valid Loss: {}\".format(val_epoch_loss/valid_loader.num_batches), \n",
+    "          \"Valid Loss: {}\".format(val_epoch_loss.mean), \n",
     "          \"Valid Accuracy:\", val_acc.value)"
    ]
   },
@@ -448,7 +448,7 @@
    "source": [
     "acc = Accuracy()\n",
     "\n",
-    "epoch_loss = 0\n",
+    "epoch_loss = Accumulator()\n",
     "start = time.time()\n",
     "model.eval()\n",
     "for batch in test_loader:\n",
@@ -456,9 +456,9 @@
     "    out = model(batch)\n",
     "    loss_val = loss(out, labels)\n",
     "    acc.update(out.argmax(dim=1), labels)\n",
-    "    epoch_loss += loss_val.item()\n",
+    "    epoch_loss.update(loss_val.item())\n",
     "end = time.time()\n",
-    "print(\"Loss: {}, Accuracy: {}\".format(epoch_loss/test_loader.num_batches, acc.value), \"Time:\", end-start)"
+    "print(\"Loss: {}, Accuracy: {}\".format(epoch_loss.mean, acc.value), \"Time:\", end-start)"
    ]
   },
   {