diff --git a/section_2_sensor_data_processing/6_grid_mapping.ipynb b/section_2_sensor_data_processing/6_grid_mapping.ipynb
index da67918..fe1ad0c 100644
--- a/section_2_sensor_data_processing/6_grid_mapping.ipynb
+++ b/section_2_sensor_data_processing/6_grid_mapping.ipynb
@@ -515,7 +515,6 @@
     "### Train the Model\n",
     "We will train the model for 5 epochs, i.e. the model sees every training sample 5 times during training. We add two callbacks that are executed regularly during the training. \n",
     "\n",
-    "- The `tensorboard_callback` will write down the current training loss after each optmization step (i.e. after each batch). Additionally, training and validation loss will be calculated after each epoch. Both graphs are visualized in the TensorBoard window below. \n",
     "- The `model_checkpoint_callback` will save the current state of the trained model parameters (weights) after each epoch to a HDF5 file. "
    ]
   },
@@ -637,7 +636,7 @@
    "source": [
     "### Test a Pretrained Model on Real-World Data\n",
     "\n",
-    "We have added a model that was trained on a dataset of 10.000 samples for 100 epochs and will use this now to predict occupancy grid maps from real-world lidar point clouds. You can see the results also in the **\"Images\" tab in TensorBoard**. Can you imagine how these occupancy grid maps can be useful for automated driving?"
+    "We have added a model that was trained on a dataset of 10.000 samples for 100 epochs and will use this now to predict occupancy grid maps from real-world lidar point clouds. You can see the results if you execute the cell below. Can you imagine how these occupancy grid maps can be useful for automated driving?"
    ]
   },
   {