diff --git a/python/tutorials/finetune-CNN-catsvsdogs.ipynb b/python/tutorials/finetune-CNN-catsvsdogs.ipynb
index 758cc5eba..44383099b 100644
--- a/python/tutorials/finetune-CNN-catsvsdogs.ipynb
+++ b/python/tutorials/finetune-CNN-catsvsdogs.ipynb
@@ -9,9 +9,9 @@
    "source": [
     "# Fine-tuning a CNN with MXNet: Cats vs Dogs (Kaggle Redux)\n",
     "\n",
-    "In this tutorial we'll learn how to build a model to classifiy if an image is a cat or a dog. We'll use a pre-trained [imagenet](http://www.image-net.org/) model from the MXNet [model zoo](http://data.mxnet.io/models/). For practical problems we may not have a large dataset, hence its difficult to train these generalized models. However we can take advantage of models that are pre-trained on large dataset like imagenet where in the model has already learnt a lot of the image features. \n",
+    "In this tutorial we'll learn how to build a model to classifiy if an image is a cat or a dog. We'll use a pre-trained [imagenet](http://www.image-net.org/) model from the MXNet [model zoo](http://data.mxnet.io/models/). For practical problems, we may not have a large dataset, hence it's difficult to train these generalized models. However we can take advantage of models that are pre-trained on large dataset like imagenet where in the model has already learnt a lot of the image features. \n",
     "\n",
-    "The model used is based on the Convolution Neural Network (CNN) architecture. CNN's consist of multiple layer of fields that are model on biological visual receptors. At each layer the neuron collection process portions of input images and the outputs get tiled so as to obtain a higher level representation of the image. For more details on the how CNN's work check out [CS231n course](http://cs231n.github.io/convolutional-networks/#overview) and [MNIST example](https://github.com/dmlc/mxnet-notebooks/blob/master/python/tutorials/mnist.ipynb) with MXNet.\n",
+    "The model used is based on the Convolution Neural Network (CNN) architecture. CNN consists of multiple layer of fields that are model on biological visual receptors. At each layer the neuron collection process portions of input images and the outputs get tiled so as to obtain a higher level representation of the image. For more details on the how CNN's work check out [CS231n course](http://cs231n.github.io/convolutional-networks/#overview) and [MNIST example](https://github.com/dmlc/mxnet-notebooks/blob/master/python/tutorials/mnist.ipynb) with MXNet.\n",
     "\n",
     "To fine-tune a network we'll update all of the network’s weights and also replace the last fully-connected layer with the new number of output classes. In most cases to train we use a smaller learning rate given that we typically may have a small dataset. For more in depth reading on fine-tuning with MXNet check this [tutorial](http://mxnet.io/how_to/finetune.html)\n"
    ]