fix part of type error

Signed-off-by: Yiheng Wang <[email protected]>

Author: yiheng-wang-nv

generation/image_to_image_translation/tutorial_segmentation_with_ddpm.ipynb

@@ -727,7 +727,7 @@
                 "    model.train()\n",
                 "    epoch_loss = 0\n",
                 "\n",
-                "    for step, data in enumerate(train_loader):\n",
+                "    for _step, data in enumerate(train_loader):\n",
                 "        images = data[\"image\"].to(device)\n",
                 "        seg = data[\"label\"].to(device)  # this is the ground truth segmentation\n",
                 "        optimizer.zero_grad(set_to_none=True)\n",
@@ -741,7 +741,8 @@
                 "            )  # we only add noise to the segmentation mask\n",
                 "            combined = torch.cat(\n",
                 "                (images, noisy_seg), dim=1\n",
-                "            )  # we concatenate the brain MR image with the noisy segmenatation mask, to condition the generation process\n",
+                "            # we concatenate the brain MR image with the noisy segmenatation mask, to condition the generation process\n",
+                "            )\n",
                 "            prediction = model(x=combined, timesteps=timesteps)\n",
                 "            # Get model prediction\n",
                 "            loss = F.mse_loss(prediction.float(), noise.float())\n",
@@ -750,11 +751,11 @@
                 "        scaler.update()\n",
                 "        epoch_loss += loss.item()\n",
                 "\n",
-                "    epoch_loss_list.append(epoch_loss / (step + 1))\n",
+                "    epoch_loss_list.append(epoch_loss / (_step + 1))\n",
                 "    if (epoch) % val_interval == 0:\n",
                 "        model.eval()\n",
                 "        val_epoch_loss = 0\n",
-                "        for step, data_val in enumerate(val_loader):\n",
+                "        for _step, data_val in enumerate(val_loader):\n",
                 "            images = data_val[\"image\"].to(device)\n",
                 "            seg = data_val[\"label\"].to(device)  # this is the ground truth segmentation\n",
                 "            timesteps = torch.randint(0, 1000, (len(images),)).to(device)\n",
@@ -766,8 +767,8 @@
                 "                    prediction = model(x=combined, timesteps=timesteps)\n",
                 "                    val_loss = F.mse_loss(prediction.float(), noise.float())\n",
                 "            val_epoch_loss += val_loss.item()\n",
-                "        print(\"Epoch\", epoch, \"Validation loss\", val_epoch_loss / (step + 1))\n",
-                "        val_epoch_loss_list.append(val_epoch_loss / (step + 1))\n",
+                "        print(\"Epoch\", epoch, \"Validation loss\", val_epoch_loss / (_step + 1))\n",
+                "        val_epoch_loss_list.append(val_epoch_loss / (_step + 1))\n",
                 "\n",
                 "torch.save(model.state_dict(), \"./segmodel.pt\")\n",
                 "total_time = time.time() - total_start\n",
@@ -1222,7 +1223,7 @@
                 "n = 5\n",
                 "input_img = inputimg[None, None, ...].to(device)\n",
                 "ensemble = []\n",
-                "for k in range(5):\n",
+                "for _ in range(5):\n",
                 "    noise = torch.randn_like(input_img).to(device)\n",
                 "    current_img = noise  # for the segmentation mask, we start from random noise.\n",
                 "    combined = torch.cat(\n",
@@ -1243,7 +1244,8 @@
                 "                    chain = torch.cat((chain, current_img.cpu()), dim=-1)\n",
                 "                combined = torch.cat(\n",
                 "                    (input_img, current_img), dim=1\n",
-                "                )  # in every step during the denoising process, the brain MR image is concatenated to add anatomical information\n",
+                "                # in every step during the denoising process, the brain MR image is concatenated to add anatomical information\n",
+                "                )\n",
                 "\n",
                 "    plt.style.use(\"default\")\n",
                 "    plt.imshow(chain[0, 0, ..., 64:].cpu(), vmin=0, vmax=1, cmap=\"gray\")\n",