3-Improving Performance.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a href=\"https://www.nvidia.com/en-us/deep-learning-ai/education/\"> <img src=\"images/DLI Header.png\" alt=\"Header\" style=\"width: 400px;\"/> </a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "raw_mimetype": "text/markdown"
   },
   "source": [
    "# Improving your Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that you've learned to successfully train a model, let's work towards a state of the art model. In this lab, we'll learn the levers that you, as a deep learning practitioner, will use to navigate towards desired results. In the process, we'll start to peel back the layers around the technology that makes this possible.\n",
    "\n",
    "Let's bring back our dog/cat classifier. \n",
    "\n",
    "## [Start from DIGITS' home screen](/digits/)\n",
    "\n",
    "Choose the model you created, \"Dogs vs. Cats\".\n",
    "\n",
    "Among other things, DIGITS will display the graph that was generated as the model was being trained. \n",
    "\n",
    "![](images/graphfromfirsttraining.png)\n",
    "\n",
    "Three quantities are reported: training loss, validation loss, and accuracy.  The values of training and validation loss should have decreased from epoch to epoch, although they may jump around some.  The accuracy is the measure of the ability of the model to correctly classify the validation data.  If you hover your mouse over any of the data points, you will see its exact value.  In this case, the accuracy at the last epoch is about 80%.  Your results might be slightly different than what is shown here since the initial networks are generated randomly.\n",
    "\n",
    "Analyzing this graph, one thing that jumps out is that accuracy is increasing over time and that loss is decreasing. A natural question may be, \"will the model keep improving if we let it train longer?\" This is the first intervention that we'll experiment with and discuss. \n",
    "\n",
    "## Study more\n",
    "\n",
    "Following the advice of parents and teachers everywhere, let's work to improve the accuracy of our model by asking it to study more. \n",
    "\n",
    "An **epoch** is one complete presentation of the data through the network. Earlier in the course, we compared one epoch through one trip through a deck of flashcards. \"Studying more\" would entail running more epochs from where our model last left off. \n",
    "\n",
    "To gain access to that *state*, scroll to the bottom of your model page and click the big green button labeled: \"Make Pretrained Model.\"\n",
    "\n",
    "![](images/pretrainedbutton.PNG)\n",
    "\n",
    "This will save two things:\n",
    "\n",
    "1. The \"network architecture\" that you chose when selecting \"AlexNet.\"\n",
    "2. What the model has \"learned\" in the form of the parameters that have been adjusted as your network worked through the data in the first 5 epochs.\n",
    "\n",
    "### Learning rate\n",
    "\n",
    "Along with the current state of the model, it is important to understand how it was changing. Take a look at the graph below the training graph:\n",
    "\n",
    "![](images/learningrate.png)\n",
    "\n",
    "You may be wondering three things:\n",
    "\n",
    "1) What is \"learning rate?\"  \n",
    "2) Why does the learning rate decrease throughout the training session?  \n",
    "3) Who controls that?  \n",
    "\n",
    "1) Learning rate is the rate at which each \"weight\" changes during training. Each weight is moving in the direction that reduces loss at a value multiplied by the learning rate.  \n",
    "2) The learning rate decreases throughout the training session because the network is getting closer to its ideal solution. At first, the network being trained knows *nothing* about what it may see. Nevermind dogs, it doesn't know whether it's about to see images or sensor data, etc. Large jumps towards the ideal solution make sense. After a few epochs, the weights are getting better so it's important for the network to *be less reactive* to each image it sees.  \n",
    "3) You control the learning rate. The learning rate is one of many \"hyperparameters\" that we set when setting up the training session.  We'll make an adjustment to it in just a minute. Here's why.  \n",
    "\n",
    "Since at the end of the session the learning rate was slow, when starting from a pretrained network, we should pick up where we left off. Let's set up a training job from this pretrained model starting at the learning rate of .0001.  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Starting from a pretrained model\n",
    "\n",
    "We can now create a new model from this starting point. Go back to DIGITS' home screen by selecting \"DIGITS\" in the top left corner and create a new Image Classification model like before. \n",
    "\n",
    "**New Model (Images) -> Classification **\n",
    "\n",
    "- Select the same dataset - (Dogs and Cats)\n",
    "- Choose some number of epochs between 3 and 8. (Note that in creating a model from scratch, this is where you could have requested more epochs originally.)\n",
    "- Change the learning rate to 0.0001  \n",
    "- Select \"advanced options\" to \"fix\" the learning rate at that value  \n",
    "- This time, instead of choosing a \"Standard Network,\" select \"Pretrained Networks.\" \n",
    "- Select the pretrained model that you just created, \"Dogs vs. Cats\".\n",
    "- Name your model - We chose \"Study more\"\n",
    "- Click Create\n",
    "\n",
    "Your settings should look like:\n",
    "\n",
    "![](images/pretrainedmodelsetup.png)\n",
    "\n",
    "When you create the model, you'll get the following graph. \n",
    "\n",
    "![](images/pretrainedgraph.PNG)\n",
    "\n",
    "Note the following:\n",
    "\n",
    "1. As expected, the accuracy starts close to where our first model left off, 80%.  \n",
    "2. Accuracy DOES continue to increase, showing that increasing the number of epochs often does increase performance.  \n",
    "3. The *rate* of increase in accuracy slows down, showing that more trips through the same data can't be the only way to increase performance.   \n",
    "\n",
    "There are four categories of levers that you can manipulate to improve performance. Time spent learning about each of them will pay off in the performance of your models.  \n",
    "\n",
    "1) **Data** - A large and diverse enough dataset to represent the environment where our model should work. Data curation is an art form in itself.    \n",
    "2) **Hyperparameters** - Making changes to options like learning rate are like changing your training \"style.\" Currently, finding the right hyperparameters is a manual process learned through experimentation. As you build intuition about what types of jobs respond well to what hyperparameters, your performance will increase.  \n",
    "3) **Training time** - More epochs improve performance to a point. At some point, too much training will result in overfitting (humans are guilty of this too), so this can not be the only intervention you apply.  \n",
    "4) **Network architecture** - We'll begin to experiment with network architecture in the next section. This is listed as the last intervention to push back against a false myth that to engage in solving problems with deep learning, people need mastery of network architecture. This field is fascinating and powerful, and improving your skills is a study in math.  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Deploying award winning models\n",
    "\n",
    "Instead of working through the deployment of the model you just retrained, it's time to introduce a shortcut to performance: deploying expert pretrained models.\n",
    "\n",
    "In this section, you'll learn to deploy other people's networks so that you can get the performance gains of their research, compute time, and data curation.\n",
    "\n",
    "Recall that the specific deep learning workflow this course started with is *image classification.* One reason we start with this task is because it is one of the most solved challenges in Deep Learning. It has benefited from the research community refining solutions to a competition called [\"ImageNet.\"](https://qz.com/1034972/the-data-that-changed-the-direction-of-ai-research-and-possibly-the-world/)\n",
    "\n",
    "\"ImageNet\" is a large dataset with 1000 classes of common images. The competition granted awards for the research teams that had the lowest loss against this dataset. The network we have been working with, AlexNet, won Imagenet in 2012. Teams from Google and Microsoft have been winners since then. \n",
    "\n",
    "Here's the exciting part. Not only can we use their network architecture, we can even use their trained weights, acquired through the manipulation of the four levers above: data, hyperparameters, training time, and network architecture. Without any training or data collection, we can *deploy* award winning neural networks.\n",
    "\n",
    "All we need to deploy one of these models are the model's architecture and weights. A quick Google search for \"pretrained model alexnet imagenet caffe\" returns multiple pages to download this model.\n",
    "\n",
    "We'll download them both using a tool called wget. Wget is a great way of downloading data from the web directly to the server you're working on without pulling it to your local machine first. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--2019-02-01 18:23:47--  http://dl.caffe.berkeleyvision.org/bvlc_alexnet.caffemodel\n",
      "Resolving dl.caffe.berkeleyvision.org (dl.caffe.berkeleyvision.org)... 169.229.222.251\n",
      "Connecting to dl.caffe.berkeleyvision.org (dl.caffe.berkeleyvision.org)|169.229.222.251|:80... connected.\n",
      "HTTP request sent, awaiting response... 200 OK\n",
      "Length: 243862414 (233M) [application/octet-stream]\n",
      "Saving to: ‘bvlc_alexnet.caffemodel’\n",
      "\n",
      "bvlc_alexnet.caffem 100%[===================>] 232.56M  16.0MB/s    in 16s     \n",
      "\n",
      "2019-02-01 18:24:03 (14.6 MB/s) - ‘bvlc_alexnet.caffemodel’ saved [243862414/243862414]\n",
      "\n",
      "--2019-02-01 18:24:03--  https://raw.githubusercontent.com/BVLC/caffe/master/models/bvlc_alexnet/deploy.prototxt\n",
      "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.200.133\n",
      "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.200.133|:443... connected.\n",
      "HTTP request sent, awaiting response... 200 OK\n",
      "Length: 3629 (3.5K) [text/plain]\n",
      "Saving to: ‘deploy.prototxt’\n",
      "\n",
      "deploy.prototxt     100%[===================>]   3.54K  --.-KB/s    in 0s      \n",
      "\n",
      "2019-02-01 18:24:03 (71.7 MB/s) - ‘deploy.prototxt’ saved [3629/3629]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "!wget http://dl.caffe.berkeleyvision.org/bvlc_alexnet.caffemodel\n",
    "!wget https://raw.githubusercontent.com/BVLC/caffe/master/models/bvlc_alexnet/deploy.prototxt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Those are the same two files that DIGITS generated when we trained a model from scratch. The only other file we took from DIGITS was the mean image that was used during training. We can download that below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--2019-02-01 18:24:11--  https://github.com/BVLC/caffe/blob/master/python/caffe/imagenet/ilsvrc_2012_mean.npy?raw=true\n",
      "Resolving github.com (github.com)... 192.30.253.112, 192.30.253.113\n",
      "Connecting to github.com (github.com)|192.30.253.112|:443... connected.\n",
      "HTTP request sent, awaiting response... 302 Found\n",
      "Location: https://github.com/BVLC/caffe/raw/master/python/caffe/imagenet/ilsvrc_2012_mean.npy [following]\n",
      "--2019-02-01 18:24:11--  https://github.com/BVLC/caffe/raw/master/python/caffe/imagenet/ilsvrc_2012_mean.npy\n",
      "Reusing existing connection to github.com:443.\n",
      "HTTP request sent, awaiting response... 302 Found\n",
      "Location: https://raw.githubusercontent.com/BVLC/caffe/master/python/caffe/imagenet/ilsvrc_2012_mean.npy [following]\n",
      "--2019-02-01 18:24:12--  https://raw.githubusercontent.com/BVLC/caffe/master/python/caffe/imagenet/ilsvrc_2012_mean.npy\n",
      "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.200.133\n",
      "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.200.133|:443... connected.\n",
      "HTTP request sent, awaiting response... 200 OK\n",
      "Length: 1572944 (1.5M) [application/octet-stream]\n",
      "Saving to: ‘ilsvrc_2012_mean.npy?raw=true’\n",
      "\n",
      "ilsvrc_2012_mean.np 100%[===================>]   1.50M  --.-KB/s    in 0.08s   \n",
      "\n",
      "2019-02-01 18:24:12 (20.0 MB/s) - ‘ilsvrc_2012_mean.npy?raw=true’ saved [1572944/1572944]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "!wget https://github.com/BVLC/caffe/blob/master/python/caffe/imagenet/ilsvrc_2012_mean.npy?raw=true\n",
    "!mv ilsvrc_2012_mean.npy?raw=true ilsvrc_2012_mean.npy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Those files are now all stored on this server as: <pre>/dli/tasks/task4/task/deploy.prototxt</pre><pre>/dli/tasks/task4/task/bvlc_alexnet.caffemodel</pre><pre>/dli/tasks/task4/task/ilsvrc_2012_mean.npy</pre>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's use what we already know about deployment to run an image through this model. Start by initializing the model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import caffe\n",
    "import numpy as np\n",
    "caffe.set_mode_gpu()\n",
    "import matplotlib.pyplot as plt #matplotlib.pyplot allows us to visualize results\n",
    "\n",
    "ARCHITECTURE = 'deploy.prototxt'\n",
    "WEIGHTS = 'bvlc_alexnet.caffemodel'\n",
    "MEAN_IMAGE = 'ilsvrc_2012_mean.npy'\n",
    "TEST_IMAGE = '/dli/data/BeagleImages/louietest2.JPG'\n",
    "\n",
    "# Initialize the Caffe model using the model trained in DIGITS\n",
    "net = caffe.Classifier(ARCHITECTURE, WEIGHTS) #Each \"channel\" of our images are 256 x 256"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then create an input the network expects. Note that this is different than the *preprocessing* used in the last model. To learn how imagenet was preprocessed, the documentation was clearly presented on the [Caffe website](http://caffe.berkeleyvision.org/gathered/examples/imagenet.html):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Load the image\n",
    "image= caffe.io.load_image(TEST_IMAGE)\n",
    "plt.imshow(image)\n",
    "plt.show()\n",
    "\n",
    "#Load the mean image\n",
    "mean_image = np.load(MEAN_IMAGE)\n",
    "mu = mean_image.mean(1).mean(1)  # average over pixels to obtain the mean (BGR) pixel values\n",
    "\n",
    "# create transformer for the input called 'data'\n",
    "transformer = caffe.io.Transformer({'data': net.blobs['data'].data.shape})\n",
    "transformer.set_transpose('data', (2,0,1))  # move image channels to outermost dimension\n",
    "transformer.set_mean('data', mu)            # subtract the dataset-mean value in each channel\n",
    "transformer.set_raw_scale('data', 255)      # rescale from [0, 1] to [0, 255]\n",
    "transformer.set_channel_swap('data', (2,1,0))  # swap channels from RGB to BGR\n",
    "# set the size of the input (we can skip this if we're happy with the default; we can also change it later, e.g., for different batch sizes)\n",
    "net.blobs['data'].reshape(1,        # batch size\n",
    "                          3,         # 3-channel (BGR) images\n",
    "                          227, 227)  # image size is 227x227\n",
    "\n",
    "transformed_image = transformer.preprocess('data', image)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Run the function and visualize the output."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'prob': array([[  2.31780262e-09,   2.58294519e-09,   3.19525961e-09,\n",
       "           2.09216755e-09,   5.00786212e-09,   2.04342987e-09,\n",
       "           2.37229258e-09,   9.16246246e-11,   4.86508278e-10,\n",
       "           5.01131137e-09,   1.35739935e-08,   9.87543114e-09,\n",
       "           3.42600948e-11,   1.11983944e-09,   3.58725938e-10,\n",
       "           7.21391349e-11,   1.98810324e-09,   3.15641522e-08,\n",
       "           3.18406670e-08,   7.21174453e-10,   8.27692492e-09,\n",
       "           1.42624259e-08,   2.83560397e-09,   3.29977219e-08,\n",
       "           3.40230094e-10,   7.50220686e-09,   9.47929624e-10,\n",
       "           1.18161825e-09,   2.00934576e-08,   2.79246071e-10,\n",
       "           1.43229650e-09,   2.25081864e-09,   8.54826698e-09,\n",
       "           1.04760878e-09,   3.35014100e-10,   1.14679100e-10,\n",
       "           8.23971502e-10,   2.29677444e-10,   1.93692991e-08,\n",
       "           3.21901672e-10,   2.40228504e-09,   1.76278014e-09,\n",
       "           2.23937793e-08,   5.04234487e-10,   4.73921236e-10,\n",
       "           5.41517942e-10,   1.59301594e-09,   2.94658253e-09,\n",
       "           3.30536420e-09,   1.88455682e-10,   2.42557197e-10,\n",
       "           3.91544575e-08,   9.13127296e-10,   7.18308080e-10,\n",
       "           3.91063715e-09,   1.42117196e-09,   2.83355472e-09,\n",
       "           8.94371688e-10,   2.96192459e-10,   3.58631702e-09,\n",
       "           1.15298526e-09,   9.82424240e-08,   9.56613988e-09,\n",
       "           6.60167165e-09,   2.98541281e-09,   3.81802989e-09,\n",
       "           6.07087047e-10,   1.03429105e-08,   1.84803317e-09,\n",
       "           2.17566673e-10,   2.10692841e-09,   2.17953056e-09,\n",
       "           2.55984345e-10,   2.13018603e-10,   3.42959411e-10,\n",
       "           2.47837528e-10,   4.29975500e-10,   5.16272858e-10,\n",
       "           8.83846241e-09,   4.32812564e-10,   1.06519922e-08,\n",
       "           5.60258417e-09,   2.87199642e-09,   1.04422400e-08,\n",
       "           6.97323355e-09,   6.76498368e-09,   1.32655469e-08,\n",
       "           5.51607160e-09,   3.71228248e-09,   1.70114181e-08,\n",
       "           6.56049312e-11,   3.24078153e-10,   1.34727718e-09,\n",
       "           5.49251267e-09,   2.62909583e-09,   2.57394384e-10,\n",
       "           1.17886056e-09,   5.02038100e-09,   6.46285514e-09,\n",
       "           4.18129993e-08,   3.78503701e-10,   2.25807852e-08,\n",
       "           1.12564502e-09,   4.14385717e-08,   1.66815994e-07,\n",
       "           1.90907681e-08,   1.44820456e-09,   4.11408407e-10,\n",
       "           1.48215121e-10,   1.29495650e-10,   2.48513743e-08,\n",
       "           6.07961903e-10,   7.35213579e-09,   1.88404758e-09,\n",
       "           2.01396833e-09,   2.97429743e-08,   1.53398549e-09,\n",
       "           1.45497059e-10,   3.33931494e-10,   4.88923346e-09,\n",
       "           9.46865786e-09,   3.21535326e-10,   1.07690780e-10,\n",
       "           1.85644042e-10,   4.88957896e-09,   6.65893385e-09,\n",
       "           3.48332974e-10,   6.54490107e-09,   1.05513676e-09,\n",
       "           7.96083921e-10,   4.10285445e-10,   8.18345613e-10,\n",
       "           9.83660708e-10,   4.72020423e-10,   3.70803157e-08,\n",
       "           2.57368848e-10,   2.94776452e-08,   4.09844558e-09,\n",
       "           3.95256317e-09,   3.61656589e-07,   8.05931251e-08,\n",
       "           2.81567369e-09,   3.52093177e-09,   3.89943011e-08,\n",
       "           2.02161177e-09,   4.69780055e-08,   4.12341166e-07,\n",
       "           2.01119676e-09,   2.31566810e-09,   1.01574327e-09,\n",
       "           5.23497818e-07,   7.16141949e-04,   9.43874028e-08,\n",
       "           1.43127581e-08,   6.54280825e-08,   8.63964615e-08,\n",
       "           5.75015690e-07,   1.75384332e-06,   6.87049876e-04,\n",
       "           1.82276294e-02,   5.31515070e-07,   1.07787130e-03,\n",
       "           4.33581322e-01,   7.21495831e-04,   2.40483228e-03,\n",
       "           8.78719584e-05,   1.52991265e-01,   2.57082045e-01,\n",
       "           2.07484458e-02,   1.55813832e-05,   1.50054950e-06,\n",
       "           1.18360175e-02,   5.19200377e-02,   2.89231748e-03,\n",
       "           2.93396454e-07,   3.20774433e-07,   1.84856672e-02,\n",
       "           1.88944213e-07,   7.24275815e-05,   2.20991299e-03,\n",
       "           1.37995998e-03,   6.79287723e-06,   5.50261648e-05,\n",
       "           8.62221583e-09,   2.09215195e-05,   2.42244209e-06,\n",
       "           1.18097716e-07,   3.07294634e-08,   2.98748382e-05,\n",
       "           8.38789947e-06,   4.27471406e-07,   1.10918484e-06,\n",
       "           3.38167752e-07,   2.70218607e-07,   1.55805935e-07,\n",
       "           7.85800296e-07,   8.99289205e-07,   1.29322700e-07,\n",
       "           8.85969129e-08,   3.65486130e-09,   1.00482815e-07,\n",
       "           1.05782139e-08,   1.87839817e-07,   2.17473950e-08,\n",
       "           1.15856098e-08,   7.64707875e-06,   3.35500113e-06,\n",
       "           8.52533092e-04,   5.91118075e-03,   6.54539617e-05,\n",
       "           7.56352310e-05,   1.53286484e-04,   3.43769261e-06,\n",
       "           6.09310973e-06,   6.30253851e-07,   5.56956977e-04,\n",
       "           2.79243295e-07,   2.79372756e-07,   6.75759156e-06,\n",
       "           3.45881426e-06,   9.73921306e-06,   1.10547704e-08,\n",
       "           5.82830853e-06,   2.51280117e-08,   3.64565023e-09,\n",
       "           1.88537433e-05,   6.68615274e-10,   8.19354027e-04,\n",
       "           1.93915062e-09,   2.62077098e-08,   1.92937136e-06,\n",
       "           5.96361633e-06,   3.02838071e-05,   1.09202025e-09,\n",
       "           2.46637660e-06,   1.49047692e-05,   1.44370670e-05,\n",
       "           3.97284239e-05,   1.06102275e-03,   1.71118685e-07,\n",
       "           4.09935368e-03,   5.96175145e-04,   9.46131258e-05,\n",
       "           8.08758195e-05,   2.08251618e-06,   2.48623269e-06,\n",
       "           2.09723832e-03,   9.43254963e-06,   5.82914035e-05,\n",
       "           9.14920861e-07,   9.62616105e-06,   2.86219201e-05,\n",
       "           7.97322264e-10,   1.23093824e-03,   4.56247250e-07,\n",
       "           3.15688524e-08,   4.51060664e-08,   2.16271269e-06,\n",
       "           2.41284703e-09,   1.58735205e-07,   2.21433893e-07,\n",
       "           1.85853055e-10,   2.02043157e-06,   1.93700413e-04,\n",
       "           8.33063677e-04,   5.22957407e-07,   3.29161452e-07,\n",
       "           9.62242575e-06,   1.84352702e-05,   1.54617027e-07,\n",
       "           2.83806187e-08,   2.08711370e-07,   1.78517411e-07,\n",
       "           3.30742681e-03,   1.32192633e-06,   4.36303708e-06,\n",
       "           1.13358840e-06,   9.23338845e-08,   9.89056375e-08,\n",
       "           1.84910594e-08,   2.21943495e-08,   6.91018398e-09,\n",
       "           1.91189358e-08,   1.79916929e-10,   3.99101737e-07,\n",
       "           7.53470673e-08,   4.94820961e-06,   2.00577999e-09,\n",
       "           1.51601647e-08,   7.27286897e-10,   1.60621937e-07,\n",
       "           8.68287486e-07,   7.23802662e-09,   9.20833463e-06,\n",
       "           5.22570531e-07,   2.30953834e-08,   7.16372668e-08,\n",
       "           6.41035669e-10,   3.58363188e-08,   2.00201342e-07,\n",
       "           5.45833112e-09,   2.48082284e-08,   5.86846838e-09,\n",
       "           9.09173681e-10,   2.99404634e-09,   3.31386096e-09,\n",
       "           4.15488195e-08,   4.74219908e-09,   2.07190687e-09,\n",
       "           4.44274884e-09,   1.52995394e-08,   1.51625001e-09,\n",
       "           5.25324184e-09,   1.95770511e-10,   6.05466954e-10,\n",
       "           4.80067985e-09,   2.30736069e-10,   1.45399193e-09,\n",
       "           6.05573203e-10,   8.03360134e-09,   3.78399934e-09,\n",
       "           5.42680478e-09,   5.37410394e-10,   3.13767817e-10,\n",
       "           7.49806939e-10,   5.07321518e-10,   3.50855456e-10,\n",
       "           1.94622429e-09,   5.98196215e-10,   7.39223460e-10,\n",
       "           3.52298746e-09,   6.24311340e-08,   1.56078361e-09,\n",
       "           3.67303308e-08,   4.49703538e-11,   6.43391473e-09,\n",
       "           1.86874760e-09,   1.76450221e-09,   5.12760074e-08,\n",
       "           4.27188638e-07,   1.67252381e-08,   2.19561315e-07,\n",
       "           1.09662366e-07,   2.36013520e-08,   1.59437485e-09,\n",
       "           2.12180476e-06,   1.20645927e-06,   1.41411540e-08,\n",
       "           3.72456707e-05,   1.48202498e-05,   5.14147018e-07,\n",
       "           4.95808422e-07,   2.96624989e-06,   6.22972084e-06,\n",
       "           1.14241834e-06,   3.03919279e-07,   1.32059324e-06,\n",
       "           2.66574620e-08,   5.44561622e-08,   4.66554894e-07,\n",
       "           2.11406572e-07,   2.04882711e-09,   9.90278171e-09,\n",
       "           1.42756462e-09,   1.20546828e-09,   3.90156851e-10,\n",
       "           3.55922269e-09,   1.28198629e-09,   3.79351164e-08,\n",
       "           6.61401933e-10,   8.95412100e-08,   1.05103027e-07,\n",
       "           1.75719155e-08,   1.38849767e-08,   6.94581246e-08,\n",
       "           3.14304693e-09,   7.33340073e-08,   4.68179273e-10,\n",
       "           1.19719941e-08,   1.09361107e-08,   7.77662290e-09,\n",
       "           7.05395165e-09,   5.05683531e-08,   7.25336378e-08,\n",
       "           1.52486095e-07,   4.29060876e-08,   3.73686468e-08,\n",
       "           1.54898689e-08,   6.29215720e-08,   3.53566287e-09,\n",
       "           1.10925935e-09,   4.22507762e-09,   9.07100339e-10,\n",
       "           5.14725207e-10,   4.49114212e-09,   2.16667306e-09,\n",
       "           1.42031234e-10,   7.69821451e-09,   8.12477996e-09,\n",
       "           5.02475883e-09,   4.40474004e-08,   1.46742352e-09,\n",
       "           3.82668924e-10,   1.30848610e-09,   5.71219960e-10,\n",
       "           2.99360114e-09,   1.70700099e-09,   3.23467608e-09,\n",
       "           7.62899854e-10,   2.18218021e-09,   4.82916729e-09,\n",
       "           1.89012059e-08,   1.07714921e-08,   3.19884474e-09,\n",
       "           2.27933779e-08,   1.16983001e-08,   2.90298807e-08,\n",
       "           1.38788856e-08,   7.92805110e-09,   1.39190291e-07,\n",
       "           2.00711159e-09,   8.66144578e-09,   2.24304046e-07,\n",
       "           5.39672342e-08,   5.37015188e-09,   4.44658727e-10,\n",
       "           6.52596366e-10,   3.93903887e-09,   1.12637736e-07,\n",
       "           2.23531860e-09,   1.72427861e-09,   2.68349787e-09,\n",
       "           4.00934352e-09,   9.46030596e-08,   2.88441061e-06,\n",
       "           5.49122436e-09,   6.24290264e-09,   5.10921527e-09,\n",
       "           6.21569418e-09,   2.42414283e-10,   8.89031693e-09,\n",
       "           2.73322087e-10,   3.45379405e-08,   3.16988746e-08,\n",
       "           1.12490302e-08,   6.15705318e-08,   9.90848181e-10,\n",
       "           1.20346806e-06,   1.92348839e-08,   1.31315647e-09,\n",
       "           4.65357058e-10,   8.19106649e-09,   2.73311951e-08,\n",
       "           3.51382057e-09,   7.91242794e-10,   4.95146413e-09,\n",
       "           1.42373591e-09,   2.07219742e-07,   1.09883769e-09,\n",
       "           4.79422191e-08,   6.82403911e-09,   4.68588635e-09,\n",
       "           1.62995684e-09,   1.18476990e-07,   3.29848220e-08,\n",
       "           1.12638228e-08,   7.22781446e-10,   2.28078934e-09,\n",
       "           1.27979005e-09,   1.31697222e-08,   4.75269379e-09,\n",
       "           6.31467572e-08,   3.13633941e-09,   1.74950330e-08,\n",
       "           1.92608400e-08,   2.95456859e-09,   5.92281513e-09,\n",
       "           2.01217620e-08,   9.89532367e-09,   6.39834008e-09,\n",
       "           1.24862087e-09,   1.29255939e-08,   1.00345829e-08,\n",
       "           4.46872095e-09,   4.12474638e-10,   1.20704033e-08,\n",
       "           1.67561709e-09,   1.46346011e-07,   1.66687173e-07,\n",
       "           9.28491595e-09,   5.32116795e-10,   6.84663348e-09,\n",
       "           1.30773259e-09,   9.09093856e-10,   3.94317574e-08,\n",
       "           1.94107108e-09,   6.35626529e-09,   3.88719013e-09,\n",
       "           1.22491723e-08,   3.96027566e-09,   2.23017427e-09,\n",
       "           7.52321405e-09,   1.02346206e-07,   6.47198384e-09,\n",
       "           2.27160815e-08,   2.90425657e-08,   4.02111455e-09,\n",
       "           1.50629098e-09,   6.51017773e-09,   3.60686925e-09,\n",
       "           9.69369585e-10,   5.17730570e-09,   2.23237571e-08,\n",
       "           3.57470720e-09,   1.90944096e-08,   2.12361033e-06,\n",
       "           1.68936165e-09,   5.77494141e-09,   9.79192394e-10,\n",
       "           2.53888874e-08,   1.39026712e-09,   2.22209828e-09,\n",
       "           5.09682252e-08,   6.42151150e-08,   3.36269657e-09,\n",
       "           1.54392665e-10,   1.26459287e-09,   1.33753799e-08,\n",
       "           3.98348021e-09,   4.46882567e-08,   1.29133220e-08,\n",
       "           5.85727467e-09,   1.92029104e-09,   2.59673683e-09,\n",
       "           2.96694136e-09,   2.32296737e-09,   1.17630905e-09,\n",
       "           6.73022532e-06,   4.99061015e-10,   2.63427864e-07,\n",
       "           4.16516877e-10,   9.68477387e-09,   1.49250727e-08,\n",
       "           9.65306845e-06,   2.71079359e-09,   4.07112966e-09,\n",
       "           4.72293582e-09,   8.00927102e-10,   1.37546827e-08,\n",
       "           2.16125384e-09,   2.22222649e-08,   2.66275180e-09,\n",
       "           1.04698259e-07,   5.48152412e-09,   9.32628419e-10,\n",
       "           1.54297589e-10,   9.75766135e-09,   4.20093258e-08,\n",
       "           2.11434461e-08,   3.95018196e-09,   2.21011192e-08,\n",
       "           7.35374484e-10,   7.39246397e-09,   1.03419384e-09,\n",
       "           8.48301818e-10,   9.76782433e-10,   7.42453654e-10,\n",
       "           1.21012471e-08,   8.16534396e-09,   1.25627719e-09,\n",
       "           1.96073117e-07,   3.19287130e-09,   6.97769087e-09,\n",
       "           6.62379096e-10,   2.54582666e-09,   8.30290414e-09,\n",
       "           6.02942030e-10,   1.34998146e-08,   1.06837390e-07,\n",
       "           4.74354644e-09,   3.44748535e-10,   2.87922935e-10,\n",
       "           4.44617410e-09,   4.86184781e-09,   1.18228387e-08,\n",
       "           1.09078426e-08,   1.06936160e-09,   9.99121141e-09,\n",
       "           2.14006786e-08,   4.61131471e-08,   2.54860204e-08,\n",
       "           4.62650895e-09,   2.33583419e-09,   1.33058640e-08,\n",
       "           1.15716459e-09,   2.28660846e-09,   2.53246117e-08,\n",
       "           1.39957184e-08,   1.42112384e-08,   2.68668963e-08,\n",
       "           2.63653934e-08,   9.98604843e-09,   7.27378549e-08,\n",
       "           1.17933574e-08,   2.92283261e-07,   1.30245850e-08,\n",
       "           1.34861423e-07,   1.73008663e-09,   4.03737843e-08,\n",
       "           9.69828662e-09,   3.03689056e-08,   1.23446320e-08,\n",
       "           7.06656622e-10,   1.29569155e-08,   1.01488205e-08,\n",
       "           6.31534780e-09,   7.76707310e-10,   8.81987017e-09,\n",
       "           6.00910539e-08,   4.26523838e-09,   1.64631100e-08,\n",
       "           1.99309298e-08,   2.69553357e-08,   3.61622421e-09,\n",
       "           4.50902760e-10,   3.72415265e-10,   3.83717902e-09,\n",
       "           1.28942901e-09,   3.59733160e-10,   2.97261127e-09,\n",
       "           1.38564547e-08,   8.46264303e-09,   3.10383648e-08,\n",
       "           5.06196933e-08,   7.84182830e-10,   6.59586608e-10,\n",
       "           1.79266610e-08,   4.08804635e-09,   5.04247986e-08,\n",
       "           4.25532711e-08,   2.63737643e-10,   1.67984282e-07,\n",
       "           8.96927310e-10,   1.54178522e-07,   5.00836839e-09,\n",
       "           7.60742136e-10,   5.35900924e-09,   5.31593614e-09,\n",
       "           1.47505641e-09,   2.81668888e-09,   1.52195085e-07,\n",
       "           2.79830026e-09,   6.09856221e-10,   3.44065554e-09,\n",
       "           7.45187911e-09,   3.08041663e-08,   2.00844807e-09,\n",
       "           3.34903771e-09,   5.56279200e-09,   2.56436028e-09,\n",
       "           4.83798779e-10,   3.59879504e-09,   5.80614223e-09,\n",
       "           5.08465847e-09,   8.53943494e-09,   9.54223012e-09,\n",
       "           1.15594974e-08,   9.40465910e-08,   5.65114844e-09,\n",
       "           1.84841642e-10,   5.17923171e-09,   2.53436361e-09,\n",
       "           7.09407866e-09,   3.70122706e-08,   7.01469416e-09,\n",
       "           1.67369740e-09,   6.54736868e-05,   4.21438928e-09,\n",
       "           2.28882527e-07,   8.42191650e-10,   7.90083945e-08,\n",
       "           1.84045703e-08,   3.11481063e-10,   2.91323721e-09,\n",
       "           3.70837085e-08,   1.54866633e-10,   3.82305032e-09,\n",
       "           3.02560754e-10,   2.49476162e-10,   1.57892632e-08,\n",
       "           2.44798230e-07,   2.11274141e-07,   1.94347183e-09,\n",
       "           1.51294586e-08,   3.27330912e-10,   9.21269550e-09,\n",
       "           4.53473170e-09,   3.25500671e-09,   1.28981770e-09,\n",
       "           1.91380467e-09,   2.06214722e-07,   1.25137296e-08,\n",
       "           1.42479408e-07,   6.60687416e-09,   2.77168244e-09,\n",
       "           2.91124302e-09,   3.96933320e-10,   3.41778250e-09,\n",
       "           4.76522217e-08,   5.85773297e-09,   3.18990843e-08,\n",
       "           6.51934451e-10,   1.87824103e-10,   8.31440417e-10,\n",
       "           1.41212564e-09,   8.25620239e-10,   1.77881632e-09,\n",
       "           1.80031066e-08,   1.79144499e-09,   1.20700065e-07,\n",
       "           4.93124741e-09,   4.55983447e-08,   2.36617481e-08,\n",
       "           7.49629958e-09,   1.08007825e-09,   8.62890772e-08,\n",
       "           8.79930351e-09,   1.16953724e-09,   1.24191653e-08,\n",
       "           3.04725051e-10,   1.27343370e-07,   1.94483718e-08,\n",
       "           5.15195211e-08,   3.96104838e-08,   4.23659885e-10,\n",
       "           8.84490436e-09,   7.21350846e-09,   1.90704874e-09,\n",
       "           1.30631916e-09,   7.29027505e-09,   3.07623615e-08,\n",
       "           3.53339114e-09,   1.30065878e-08,   2.26888353e-09,\n",
       "           9.47848822e-09,   1.30381636e-07,   5.75045256e-09,\n",
       "           5.11474374e-09,   1.29831579e-09,   1.36177603e-09,\n",
       "           1.10603857e-07,   1.39825851e-09,   9.10315912e-09,\n",
       "           4.25759783e-09,   6.32592023e-09,   9.46917766e-10,\n",
       "           5.56696378e-09,   1.95511252e-09,   6.54063337e-09,\n",
       "           1.66393903e-08,   1.18932642e-09,   2.43238958e-08,\n",
       "           1.78698811e-09,   9.35648892e-09,   1.04597415e-08,\n",
       "           4.42565877e-08,   5.60304381e-09,   1.24102084e-09,\n",
       "           5.57561073e-08,   1.02026354e-09,   2.44172660e-09,\n",
       "           5.23845367e-09,   1.61923630e-08,   1.02393848e-07,\n",
       "           3.62699293e-09,   4.41617054e-09,   2.55796095e-09,\n",
       "           6.96811109e-10,   1.33691813e-08,   3.98073841e-09,\n",
       "           3.83493098e-10,   2.49599519e-09,   2.00272652e-08,\n",
       "           6.34485242e-10,   2.95025036e-08,   2.73826544e-07,\n",
       "           1.41825094e-08,   2.61943534e-09,   5.58142332e-09,\n",
       "           2.61958033e-09,   1.01648618e-08,   3.46932083e-09,\n",
       "           4.81660436e-08,   1.93615417e-08,   1.02928057e-08,\n",
       "           2.10593711e-08,   1.16057794e-08,   2.94992990e-08,\n",
       "           6.04828687e-10,   4.59698013e-09,   7.10444759e-10,\n",
       "           3.42434703e-09,   1.21370924e-09,   5.87292215e-10,\n",
       "           5.86788662e-09,   5.64829861e-05,   8.20135249e-09,\n",
       "           3.17286775e-09,   6.89743104e-07,   5.88593396e-09,\n",
       "           3.46198359e-09,   4.54011726e-08,   4.95292829e-09,\n",
       "           1.27405251e-08,   2.44748388e-10,   2.05062828e-10,\n",
       "           2.03423656e-09,   1.75437764e-09,   1.79414048e-08,\n",
       "           1.03716342e-08,   7.91857246e-10,   9.18062085e-11,\n",
       "           9.35452693e-09,   1.51394453e-08,   1.82634690e-08,\n",
       "           3.40789486e-09,   5.52329071e-10,   2.05691357e-08,\n",
       "           1.86300353e-09,   2.97430580e-10,   1.28311965e-08,\n",
       "           4.07663379e-08,   2.21939578e-09,   6.80221712e-10,\n",
       "           1.55682642e-07,   1.74246825e-10,   9.60391446e-08,\n",
       "           1.41741594e-07,   1.63012874e-08,   2.36048514e-09,\n",
       "           2.09428364e-09,   1.01421506e-07,   1.43663726e-07,\n",
       "           1.71947807e-08,   7.36287076e-09,   3.02521630e-09,\n",
       "           5.27116795e-09,   1.26817257e-09,   1.94938288e-08,\n",
       "           4.70474726e-09,   6.58274871e-07,   1.38657894e-08,\n",
       "           1.21070785e-04,   2.86590446e-10,   3.07946935e-09,\n",
       "           1.76298265e-08,   4.66220840e-09,   1.68449343e-09,\n",
       "           8.46110570e-10,   2.42036133e-08,   3.16619975e-09,\n",
       "           1.16595631e-08,   1.62938889e-08,   5.05043207e-09,\n",
       "           1.06440057e-09,   4.51239856e-07,   5.54359725e-10,\n",
       "           1.19214028e-09,   1.03150732e-08,   1.25768995e-08,\n",
       "           2.29149961e-08,   1.43035006e-09,   4.69109338e-08,\n",
       "           5.29175537e-10,   3.46907475e-10,   7.91330645e-09,\n",
       "           6.23769765e-08,   3.64273611e-10,   1.95784811e-09,\n",
       "           3.33934214e-09,   5.65142821e-09,   7.49306561e-10,\n",
       "           4.09026377e-08,   1.69330754e-08,   1.11490794e-09,\n",
       "           1.11570458e-07,   5.82789650e-10,   8.55922666e-10,\n",
       "           1.67351855e-09,   5.84215947e-08,   1.69644014e-08,\n",
       "           1.21444343e-09,   1.53633384e-09,   4.06693124e-09,\n",
       "           9.52024681e-10,   1.39912082e-08,   2.14476614e-09,\n",
       "           6.83995083e-09,   5.45815659e-08,   1.02108952e-08,\n",
       "           2.01925321e-09,   1.69114500e-08,   6.60713582e-08,\n",
       "           1.64542282e-08,   4.62900083e-08,   2.03370121e-09,\n",
       "           5.62882896e-08,   3.98372357e-09,   2.92520808e-09,\n",
       "           6.72166500e-09,   2.74501009e-08,   4.29129585e-08,\n",
       "           4.33741043e-08,   9.70163172e-10,   2.27733665e-10,\n",
       "           2.39521514e-10,   1.47451704e-08,   1.88971665e-08,\n",
       "           1.96994332e-09,   2.39021447e-09,   4.58119009e-09,\n",
       "           9.16413256e-09,   2.83365609e-10,   7.52338247e-10,\n",
       "           4.40543557e-09,   1.59321656e-09,   2.50845039e-10,\n",
       "           9.11683351e-09,   7.06857861e-09,   4.30004860e-07,\n",
       "           1.74467052e-09,   6.42121734e-08,   1.39848586e-08,\n",
       "           1.75973647e-09,   6.84332702e-10,   7.38009098e-10,\n",
       "           1.02457343e-09,   5.74614312e-09,   1.74485359e-09,\n",
       "           1.63354830e-09,   8.89102514e-10,   3.16383941e-09,\n",
       "           2.19129337e-09,   5.76722226e-10,   5.85071769e-10,\n",
       "           4.38478809e-09,   1.03755871e-09,   5.34484812e-08,\n",
       "           2.29038992e-08,   6.19334806e-10,   4.50705162e-09,\n",
       "           4.61923726e-08,   4.58914629e-10,   1.16998555e-09,\n",
       "           1.19662875e-08,   2.55975219e-09,   5.51481172e-08,\n",
       "           3.34089506e-10,   5.35509592e-09,   5.17955456e-10,\n",
       "           2.10799858e-08,   1.50125228e-08,   5.83575757e-11,\n",
       "           4.87324847e-09,   1.03735476e-09,   8.35556779e-09,\n",
       "           9.00073849e-10,   1.13103070e-07,   3.44919222e-08,\n",
       "           1.39825052e-09,   7.03649583e-09,   9.47259160e-10,\n",
       "           1.06990488e-08,   7.44618189e-09,   3.81153198e-09,\n",
       "           1.78574879e-08,   2.32537118e-08,   2.21020464e-08,\n",
       "           2.31418582e-08,   2.00962846e-09,   1.51297685e-09,\n",
       "           5.21386223e-09,   4.13715817e-09,   8.16785128e-09,\n",
       "           3.87228845e-08,   2.14325210e-10,   2.03813494e-10,\n",
       "           7.04438250e-08,   8.68248495e-09,   3.35417050e-09,\n",
       "           3.02721381e-09,   3.52214702e-09,   1.11867244e-08,\n",
       "           4.47059030e-08,   3.80252807e-09,   2.30513297e-08,\n",
       "           7.29049709e-09,   1.45533008e-08,   1.79645525e-08,\n",
       "           6.16855857e-07]], dtype=float32)}"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# copy the image data into the memory allocated for the net\n",
    "net.blobs['data'].data[...] = transformed_image\n",
    "\n",
    "### perform classification\n",
    "output = net.forward()\n",
    "\n",
    "output"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Work to make the output useful to a user.\n",
    "\n",
    "What you see above is an array containing the probabilities that our image belongs to each of 1000 classes. Let's work to make this useful."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "predicted class is: 162\n"
     ]
    }
   ],
   "source": [
    "output_prob = output['prob'][0]  # the output probability vector for the first image in the batch\n",
    "print 'predicted class is:', output_prob.argmax()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is closer. Take a look at imagenet's classes to see what that number corresponds to [here.](https://gist.github.com/yrevar/942d3a0ac09ec9e5eb3a) Better???? Let's add that functionality to our application so we have a useful end-to-end deployment.\n",
    "\n",
    "Again using wget to get a dictionary (dict) of class to label."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--2019-02-01 18:27:06--  https://raw.githubusercontent.com/HoldenCaulfieldRye/caffe/master/data/ilsvrc12/synset_words.txt\n",
      "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.200.133\n",
      "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.200.133|:443... connected.\n",
      "HTTP request sent, awaiting response... 200 OK\n",
      "Length: 31675 (31K) [text/plain]\n",
      "Saving to: ‘synset_words.txt’\n",
      "\n",
      "synset_words.txt    100%[===================>]  30.93K  --.-KB/s    in 0.01s   \n",
      "\n",
      "2019-02-01 18:27:06 (2.89 MB/s) - ‘synset_words.txt’ saved [31675/31675]\n",
      "\n",
      "output label: n02088364 beagle\n"
     ]
    }
   ],
   "source": [
    "!wget https://raw.githubusercontent.com/HoldenCaulfieldRye/caffe/master/data/ilsvrc12/synset_words.txt\n",
    "labels_file = 'synset_words.txt'\n",
    "labels = np.loadtxt(labels_file, str, delimiter='\\t')\n",
    "\n",
    "print 'output label:', labels[output_prob.argmax()]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To get a clean view of what our application does, here is the input and output of our application."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input image:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Output label:n02088364 beagle\n"
     ]
    }
   ],
   "source": [
    "print (\"Input image:\")\n",
    "plt.imshow(image)\n",
    "plt.show()\n",
    "\n",
    "print(\"Output label:\" + labels[output_prob.argmax()])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a href=\"https://www.nvidia.com/en-us/deep-learning-ai/education/\"> <img src=\"images/DLI Header.png\" alt=\"Header\" style=\"width: 400px;\"/> </a>"
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Raw Cell Format",
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}