Tensorflow basics

TF - graph exemple.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# TensorFlow basics\n",
    "\n",
    "This notebook present what is the design of TensorFlow. We we see what is a variable, an operation, and how \"graphs\" encapsulate everything and how to \"flow\" data inside a graph.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from tfboardviewer import show_graph\n",
    "import matplotlib.pyplot as plt\n",
    "import PIL\n",
    "import numpy as np\n",
    "import math"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## TensorFlow ? Graph ?\n",
    "\n",
    "TensorFlow is a framework that is able to make operation on complex variables (numbers, matrix...) and to prepare a \"graph\" that is, mainly, a definition of the entire needed object to make calculation.\n",
    "\n",
    "A graph **doesn't do anyting while we didn't start a session** to \"flow\" values inside.\n",
    "\n",
    "Variables are tensors, that means \"several variables grouped in one object\" that have \"shape\" (dimension). This is more or less \"matrices\".\n",
    "\n",
    "Tensorflow allow you to prepare calculation and is able to launch the \"flow\" on several devices, as CPU or GPU.\n",
    "\n",
    "The name TensorFlow is now obvious, we flow values in tensors.\n",
    "\n",
    "\n",
    "## Generate a graph\n",
    "\n",
    "TensorFlow proposes several types to be inserted in a graph. We can create:\n",
    "\n",
    "- variables that can be initialized and assigned inside the graph\n",
    "- placeholders that are variable we can set as \"input\" (when we will start and run a session)\n",
    "- operations that can be addition, multiplication, or more complex as matmul, sigmoid, and so on...\n",
    "\n",
    "It's **very important** to understand that this variables are **not** python variables, they are **Tensors** injected in the default graph.\n",
    "\n",
    "Something intersting is that we can use **operators** to create some operations.\n",
    "\n",
    "For example, in the following lines, we are doing:\n",
    "\n",
    "```python\n",
    "v3 = v1 + v2\n",
    "```\n",
    "\n",
    "The result is not \"8.0\" but a `tensorflow.python.framework.ops.Tensor` that will execute the \"addition\" only inside a `tf.Session`.\n",
    "\n",
    "This is a simple basic example:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "v1 and v2: <tf.Variable 'var1:0' shape=() dtype=float32_ref> <tf.Variable 'var2:0' shape=() dtype=float32_ref>\n",
      "Type of 'addop': <class 'tensorflow.python.framework.ops.Tensor'>\n",
      "It is defined like that: Tensor(\"add:0\", shape=(), dtype=float32)\n",
      "It's an addition of 'v1' and 'v2'\n",
      "Output op Tensor(\"Multiply:0\", dtype=float32)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "        <iframe seamless style=\"width:800px;height:620px;border:0\" srcdoc=\"\n",
       "        <script>\n",
       "          function load() {\n",
       "            document.getElementById(&quot;graph0.29436165906015443&quot;).pbtxt = 'node {\\n  name: &quot;var1/initial_value&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_FLOAT\\n        tensor_shape {\\n        }\\n        float_val: 4.0\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;var1&quot;\\n  op: &quot;VariableV2&quot;\\n  attr {\\n    key: &quot;container&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;shape&quot;\\n    value {\\n      shape {\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;shared_name&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;var1/Assign&quot;\\n  op: &quot;Assign&quot;\\n  input: &quot;var1&quot;\\n  input: &quot;var1/initial_value&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@var1&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;use_locking&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n  attr {\\n    key: &quot;validate_shape&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;var1/read&quot;\\n  op: &quot;Identity&quot;\\n  input: &quot;var1&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@var1&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;var2/initial_value&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_FLOAT\\n        tensor_shape {\\n        }\\n        float_val: 4.0\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;var2&quot;\\n  op: &quot;VariableV2&quot;\\n  attr {\\n    key: &quot;container&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;shape&quot;\\n    value {\\n      shape {\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;shared_name&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;var2/Assign&quot;\\n  op: &quot;Assign&quot;\\n  input: &quot;var2&quot;\\n  input: &quot;var2/initial_value&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@var2&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;use_locking&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n  attr {\\n    key: &quot;validate_shape&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;var2/read&quot;\\n  op: &quot;Identity&quot;\\n  input: &quot;var2&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@var2&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;add&quot;\\n  op: &quot;Add&quot;\\n  input: &quot;var1/read&quot;\\n  input: &quot;var2/read&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;input&quot;\\n  op: &quot;Placeholder&quot;\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;shape&quot;\\n    value {\\n      shape {\\n        unknown_rank: true\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;Multiply&quot;\\n  op: &quot;Mul&quot;\\n  input: &quot;add&quot;\\n  input: &quot;input&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n}\\n';\n",
       "          }\n",
       "        </script>\n",
       "        <link rel=&quot;import&quot; href=&quot;https://tensorboard.appspot.com/tf-graph-basic.build.html&quot; onload=load()>\n",
       "        <div style=&quot;height:600px&quot;>\n",
       "          <tf-graph-basic id=&quot;graph0.29436165906015443&quot;></tf-graph-basic>\n",
       "        </div>\n",
       "    \"></iframe>\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "tf.reset_default_graph()\n",
    "\n",
    "# this variable is at the root of graph\n",
    "v1 = tf.Variable(4., name=\"var1\")\n",
    "v2 = tf.Variable(4., name=\"var2\")\n",
    "\n",
    "print(\"v1 and v2:\", v1, v2)\n",
    "\n",
    "# create an operation using python operator overrides.\n",
    "# This will create an \"add\" operation in the graph\n",
    "addop = v1 + v2\n",
    "\n",
    "# what is the type of v3 ?\n",
    "print(\"Type of 'addop':\", type(addop))\n",
    "print(\"It is defined like that:\", addop)\n",
    "print(\"It's an addition of 'v1' and 'v2'\")\n",
    "    \n",
    "# here, we create a placeholder to be able to\n",
    "# assig a variable when we will evaluate the graph\n",
    "input_placeholder = tf.placeholder(tf.float32, name='input')\n",
    "\n",
    "# create a multiplicatoin without using\n",
    "# python operator. Also, we define a custom name\n",
    "output_op = tf.multiply(addop, input_placeholder, name=\"Multiply\")\n",
    "\n",
    "print(\"Output op\", output_op)\n",
    "\n",
    "# this is the same as making\n",
    "# output_op = addop * input_placholder\n",
    "# but without the ability to change the name\n",
    "\n",
    "# we can display the graph\n",
    "with tf.Session() as sess:\n",
    "    show_graph(sess.graph.as_graph_def())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Variables can be set in the root of the graph, or inside \"scopes\" that are \"namespaces\". Scope are only there to make graph easy to read, they don't impact how the graph is evaluated. Note nevertheless that scope impacts how we set and get values when we will call the session run."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "        <iframe seamless style=\"width:800px;height:620px;border:0\" srcdoc=\"\n",
       "        <script>\n",
       "          function load() {\n",
       "            document.getElementById(&quot;graph0.2647211261921991&quot;).pbtxt = 'node {\\n  name: &quot;var1/initial_value&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_FLOAT\\n        tensor_shape {\\n        }\\n        float_val: 4.0\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;var1&quot;\\n  op: &quot;VariableV2&quot;\\n  attr {\\n    key: &quot;container&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;shape&quot;\\n    value {\\n      shape {\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;shared_name&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;var1/Assign&quot;\\n  op: &quot;Assign&quot;\\n  input: &quot;var1&quot;\\n  input: &quot;var1/initial_value&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@var1&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;use_locking&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n  attr {\\n    key: &quot;validate_shape&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;var1/read&quot;\\n  op: &quot;Identity&quot;\\n  input: &quot;var1&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@var1&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;MyAdditionBloc/var2/initial_value&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_FLOAT\\n        tensor_shape {\\n        }\\n        float_val: 4.0\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;MyAdditionBloc/var2&quot;\\n  op: &quot;VariableV2&quot;\\n  attr {\\n    key: &quot;container&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;shape&quot;\\n    value {\\n      shape {\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;shared_name&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;MyAdditionBloc/var2/Assign&quot;\\n  op: &quot;Assign&quot;\\n  input: &quot;MyAdditionBloc/var2&quot;\\n  input: &quot;MyAdditionBloc/var2/initial_value&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@MyAdditionBloc/var2&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;use_locking&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n  attr {\\n    key: &quot;validate_shape&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;MyAdditionBloc/var2/read&quot;\\n  op: &quot;Identity&quot;\\n  input: &quot;MyAdditionBloc/var2&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@MyAdditionBloc/var2&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;MyAdditionBloc/add&quot;\\n  op: &quot;Add&quot;\\n  input: &quot;var1/read&quot;\\n  input: &quot;MyAdditionBloc/var2/read&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;input&quot;\\n  op: &quot;Placeholder&quot;\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;shape&quot;\\n    value {\\n      shape {\\n        unknown_rank: true\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;MyOps/Multiply&quot;\\n  op: &quot;Mul&quot;\\n  input: &quot;MyAdditionBloc/add&quot;\\n  input: &quot;input&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n}\\n';\n",
       "          }\n",
       "        </script>\n",
       "        <link rel=&quot;import&quot; href=&quot;https://tensorboard.appspot.com/tf-graph-basic.build.html&quot; onload=load()>\n",
       "        <div style=&quot;height:600px&quot;>\n",
       "          <tf-graph-basic id=&quot;graph0.2647211261921991&quot;></tf-graph-basic>\n",
       "        </div>\n",
       "    \"></iframe>\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "tf.reset_default_graph()\n",
    "\n",
    "# this variable is at the root of graph\n",
    "v1 = tf.Variable(4., name=\"var1\")\n",
    "\n",
    "# we can create a scope to classify variables\n",
    "with tf.variable_scope(\"MyAdditionBloc\"):\n",
    "    v2 = tf.Variable(4., name=\"var2\")\n",
    "    addop = v1 + v2\n",
    "    \n",
    "# Input placeholder will be injeted at the root\n",
    "input_placeholder = tf.placeholder(tf.float32, name='input')\n",
    "\n",
    "# we can also create a scope for operations.\n",
    "with tf.name_scope(\"MyOps\"):\n",
    "    v4 = tf.multiply(addop, input_placeholder, name=\"Multiply\")\n",
    "\n",
    "# we can display the graph\n",
    "with tf.Session() as sess:\n",
    "    show_graph(sess.graph.as_graph_def())\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you can see, the graph has got variables, scopes and operations.\n",
    "\n",
    "It's now time to try to \"run\" the graph.\n",
    "\n",
    "To be able to evaluate the graph, we need to start a `Session`. Inside the session, we need to initialize variables.\n",
    "\n",
    "Then, we can call `sess.run` giving:\n",
    "\n",
    "- what to read: here we want to read output of MyOps/Multiply\n",
    "- `fed_dict` as second argument to present values on certain placeholders, here we send \"4.\" to the \"input\" placeholder\n",
    "\n",
    "The needed syntax is \"`<tensor name>:<index>`\"."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Result of the graph operations 16.0\n"
     ]
    }
   ],
   "source": [
    "# we can use Session to launch graph and send input value\n",
    "with tf.Session() as sess:\n",
    "    initfunc = tf.global_variables_initializer()\n",
    "    sess.run(initfunc)\n",
    "    # we want to read the result of operation \"Multiply\" inside the \"MyOps\" scope\n",
    "    result = sess.run('MyOps/Multiply:0', {\n",
    "        'input:0': 2.  # and we set the value of the \"input\" placeholder\n",
    "    })\n",
    "    print(\"Result of the graph operations\", result)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And that's ok, operations are launched inside the session, it adds var1 + var2 (4+4) then multiply the result to the input (we send 2.), so (4+4) * 2 = 16.\n",
    "\n",
    "This is how we can summarize how TensorFlow works:\n",
    "\n",
    "- add vars, constants, placholder, operation\n",
    "- define the \"flow\" of values (connection between vars, operations, and so on...)\n",
    "- open a session and proceed the flow\n",
    "\n",
    "Something interesting is that the graph can store the results, variables assignations, and to be shared with \"trained\" variables. That way, TensorFlow can be trained with neural network definition and make prediction. This is the base of Machine Learining.\n",
    "\n",
    "We can now make a step further and use \"images\" as matrix to understand how to flow that kind of data. For example, using convolutions."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using images (as matrix) and create a convolutional network\n",
    "\n",
    "In the following example, we will create a new graph that is able to get an image and pass the matrix inside convolutions.\n",
    "\n",
    "Convolution is an operation that reduce a matrix making operations on cols and rows. It result of features extraction, blur, and so on...\n",
    "\n",
    "This example is **basic**, this is not a graph made to train and make image recognition, it is only an example that show you what a convolution does inside a graph.\n",
    "\n",
    "So, at first, we will create a \"input\" that gets an image sized 224x224 with 3 chanels (RGB)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "tf.reset_default_graph()\n",
    "input_image = tf.placeholder(tf.float32, shape=(None,224,224,3), name='input')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we can create a \"convolution\" - instead of making operations ourself, we can use Tensorflow subpackage that provides ready to use layers. For the example, we add an activation function that is the standard when we want to train a graph.\n",
    "\n",
    "We can tell the graph to make a certain number of convolution in \"parallel\" - for example, use 4 convolutions:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "convbloc = tf.layers.conv2d(input_image, 4, (3,3), activation='relu', name=\"conv1\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can directly use the convolution bloc output, just check the graph and open \"conv1\" block to see where."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "        <iframe seamless style=\"width:800px;height:620px;border:0\" srcdoc=\"\n",
       "        <script>\n",
       "          function load() {\n",
       "            document.getElementById(&quot;graph0.7173858200604973&quot;).pbtxt = 'node {\\n  name: &quot;input&quot;\\n  op: &quot;Placeholder&quot;\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;shape&quot;\\n    value {\\n      shape {\\n        dim {\\n          size: -1\\n        }\\n        dim {\\n          size: 224\\n        }\\n        dim {\\n          size: 224\\n        }\\n        dim {\\n          size: 3\\n        }\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/kernel/Initializer/random_uniform/shape&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv1/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_INT32\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_INT32\\n        tensor_shape {\\n          dim {\\n            size: 4\\n          }\\n        }\\n        tensor_content: &quot;\\\\003\\\\000\\\\000\\\\000\\\\003\\\\000\\\\000\\\\000\\\\003\\\\000\\\\000\\\\000\\\\004\\\\000\\\\000\\\\000&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/kernel/Initializer/random_uniform/min&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv1/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_FLOAT\\n        tensor_shape {\\n        }\\n        float_val: -0.30860671401023865\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/kernel/Initializer/random_uniform/max&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv1/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_FLOAT\\n        tensor_shape {\\n        }\\n        float_val: 0.30860671401023865\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/kernel/Initializer/random_uniform/RandomUniform&quot;\\n  op: &quot;RandomUniform&quot;\\n  input: &quot;conv1/kernel/Initializer/random_uniform/shape&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_INT32\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv1/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;seed&quot;\\n    value {\\n      i: 0\\n    }\\n  }\\n  attr {\\n    key: &quot;seed2&quot;\\n    value {\\n      i: 0\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/kernel/Initializer/random_uniform/sub&quot;\\n  op: &quot;Sub&quot;\\n  input: &quot;conv1/kernel/Initializer/random_uniform/max&quot;\\n  input: &quot;conv1/kernel/Initializer/random_uniform/min&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv1/kernel&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/kernel/Initializer/random_uniform/mul&quot;\\n  op: &quot;Mul&quot;\\n  input: &quot;conv1/kernel/Initializer/random_uniform/RandomUniform&quot;\\n  input: &quot;conv1/kernel/Initializer/random_uniform/sub&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv1/kernel&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/kernel/Initializer/random_uniform&quot;\\n  op: &quot;Add&quot;\\n  input: &quot;conv1/kernel/Initializer/random_uniform/mul&quot;\\n  input: &quot;conv1/kernel/Initializer/random_uniform/min&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv1/kernel&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/kernel&quot;\\n  op: &quot;VariableV2&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv1/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;container&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;shape&quot;\\n    value {\\n      shape {\\n        dim {\\n          size: 3\\n        }\\n        dim {\\n          size: 3\\n        }\\n        dim {\\n          size: 3\\n        }\\n        dim {\\n          size: 4\\n        }\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;shared_name&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/kernel/Assign&quot;\\n  op: &quot;Assign&quot;\\n  input: &quot;conv1/kernel&quot;\\n  input: &quot;conv1/kernel/Initializer/random_uniform&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv1/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;use_locking&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n  attr {\\n    key: &quot;validate_shape&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/kernel/read&quot;\\n  op: &quot;Identity&quot;\\n  input: &quot;conv1/kernel&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv1/kernel&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/bias/Initializer/zeros&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv1/bias&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_FLOAT\\n        tensor_shape {\\n          dim {\\n            size: 4\\n          }\\n        }\\n        float_val: 0.0\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/bias&quot;\\n  op: &quot;VariableV2&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv1/bias&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;container&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;shape&quot;\\n    value {\\n      shape {\\n        dim {\\n          size: 4\\n        }\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;shared_name&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/bias/Assign&quot;\\n  op: &quot;Assign&quot;\\n  input: &quot;conv1/bias&quot;\\n  input: &quot;conv1/bias/Initializer/zeros&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv1/bias&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;use_locking&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n  attr {\\n    key: &quot;validate_shape&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/bias/read&quot;\\n  op: &quot;Identity&quot;\\n  input: &quot;conv1/bias&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv1/bias&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/dilation_rate&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_INT32\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_INT32\\n        tensor_shape {\\n          dim {\\n            size: 2\\n          }\\n        }\\n        tensor_content: &quot;\\\\001\\\\000\\\\000\\\\000\\\\001\\\\000\\\\000\\\\000&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/Conv2D&quot;\\n  op: &quot;Conv2D&quot;\\n  input: &quot;input&quot;\\n  input: &quot;conv1/kernel/read&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;data_format&quot;\\n    value {\\n      s: &quot;NHWC&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;dilations&quot;\\n    value {\\n      list {\\n        i: 1\\n        i: 1\\n        i: 1\\n        i: 1\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;padding&quot;\\n    value {\\n      s: &quot;VALID&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;strides&quot;\\n    value {\\n      list {\\n        i: 1\\n        i: 1\\n        i: 1\\n        i: 1\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;use_cudnn_on_gpu&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/BiasAdd&quot;\\n  op: &quot;BiasAdd&quot;\\n  input: &quot;conv1/Conv2D&quot;\\n  input: &quot;conv1/bias/read&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;data_format&quot;\\n    value {\\n      s: &quot;NHWC&quot;\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv1/Relu&quot;\\n  op: &quot;Relu&quot;\\n  input: &quot;conv1/BiasAdd&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n}\\n';\n",
       "          }\n",
       "        </script>\n",
       "        <link rel=&quot;import&quot; href=&quot;https://tensorboard.appspot.com/tf-graph-basic.build.html&quot; onload=load()>\n",
       "        <div style=&quot;height:600px&quot;>\n",
       "          <tf-graph-basic id=&quot;graph0.7173858200604973&quot;></tf-graph-basic>\n",
       "        </div>\n",
       "    \"></iframe>\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "with tf.Session() as sess:\n",
    "    show_graph(sess.graph.as_graph_def())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the graph, we can (for example), get the result from \"Relu\" operation inside the \"conv1\" bloc. We can now open an image and make some transformation to inject it inside the \"input\" placeholder, and try to read the \"conv1/Relu\" output.\n",
    "\n",
    "Note that because we decided to make 4 convolution, the output of \"Relu\" will have 4 results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 222, 222, 4)\n"
     ]
    }
   ],
   "source": [
    "# Open an image, resize it to (224, 224) that is the shape of the graph input\n",
    "image = PIL.Image.open('data/train/burger/02821fe8e4e7bf47046e1249e840d9b5.jpeg')\n",
    "image = image.resize((224,224))\n",
    "\n",
    "# use numpy to reshape the array and to add a dimension\n",
    "image = np.array(image)\n",
    "image = np.expand_dims(image, axis=0)\n",
    "\n",
    "# open a session, initialize variables and inject the image\n",
    "with tf.Session() as sess:\n",
    "    init = tf.global_variables_initializer()\n",
    "    sess.run(init)\n",
    "    res = sess.run('conv1/Relu:0', {\n",
    "        'input:0' : image\n",
    "    })\n",
    "\n",
    "print(res.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As the shape has got an extra dimension (because we can use batches...) we can \"squeeze\" the array to get the output \"as an image\". Then, display the 4 convolution outputs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "res = np.squeeze(res)\n",
    "\n",
    "columns = 4\n",
    "rows = math.ceil(res.shape[2] / columns)\n",
    "\n",
    "plt.figure(figsize=(12, 6))\n",
    "for i in range(res.shape[2]):\n",
    "    plt.subplot(\"%d%d%d\" % (rows, columns, i+1))\n",
    "    plt.imshow(res[:,:,i])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can add some convolution and other operations. For example, in the new graph below, we will use:\n",
    "\n",
    "- 2 layers with 6 convolutions\n",
    "- 1 max pooling 2D (that will get max values from the convolutions)\n",
    "\n",
    "Also, to help the graph usage, we can add an output. It can be a tf.Variable, or for example a Reshape operation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "        <iframe seamless style=\"width:800px;height:620px;border:0\" srcdoc=\"\n",
       "        <script>\n",
       "          function load() {\n",
       "            document.getElementById(&quot;graph0.13782732954427268&quot;).pbtxt = 'node {\\n  name: &quot;input&quot;\\n  op: &quot;Placeholder&quot;\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;shape&quot;\\n    value {\\n      shape {\\n        dim {\\n          size: -1\\n        }\\n        dim {\\n          size: 224\\n        }\\n        dim {\\n          size: 224\\n        }\\n        dim {\\n          size: 3\\n        }\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/kernel/Initializer/random_uniform/shape&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_INT32\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_INT32\\n        tensor_shape {\\n          dim {\\n            size: 4\\n          }\\n        }\\n        tensor_content: &quot;\\\\003\\\\000\\\\000\\\\000\\\\003\\\\000\\\\000\\\\000\\\\003\\\\000\\\\000\\\\000\\\\006\\\\000\\\\000\\\\000&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/kernel/Initializer/random_uniform/min&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_FLOAT\\n        tensor_shape {\\n        }\\n        float_val: -0.27216553688049316\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/kernel/Initializer/random_uniform/max&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_FLOAT\\n        tensor_shape {\\n        }\\n        float_val: 0.27216553688049316\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/kernel/Initializer/random_uniform/RandomUniform&quot;\\n  op: &quot;RandomUniform&quot;\\n  input: &quot;conv2d/kernel/Initializer/random_uniform/shape&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_INT32\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;seed&quot;\\n    value {\\n      i: 0\\n    }\\n  }\\n  attr {\\n    key: &quot;seed2&quot;\\n    value {\\n      i: 0\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/kernel/Initializer/random_uniform/sub&quot;\\n  op: &quot;Sub&quot;\\n  input: &quot;conv2d/kernel/Initializer/random_uniform/max&quot;\\n  input: &quot;conv2d/kernel/Initializer/random_uniform/min&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d/kernel&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/kernel/Initializer/random_uniform/mul&quot;\\n  op: &quot;Mul&quot;\\n  input: &quot;conv2d/kernel/Initializer/random_uniform/RandomUniform&quot;\\n  input: &quot;conv2d/kernel/Initializer/random_uniform/sub&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d/kernel&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/kernel/Initializer/random_uniform&quot;\\n  op: &quot;Add&quot;\\n  input: &quot;conv2d/kernel/Initializer/random_uniform/mul&quot;\\n  input: &quot;conv2d/kernel/Initializer/random_uniform/min&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d/kernel&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/kernel&quot;\\n  op: &quot;VariableV2&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;container&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;shape&quot;\\n    value {\\n      shape {\\n        dim {\\n          size: 3\\n        }\\n        dim {\\n          size: 3\\n        }\\n        dim {\\n          size: 3\\n        }\\n        dim {\\n          size: 6\\n        }\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;shared_name&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/kernel/Assign&quot;\\n  op: &quot;Assign&quot;\\n  input: &quot;conv2d/kernel&quot;\\n  input: &quot;conv2d/kernel/Initializer/random_uniform&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;use_locking&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n  attr {\\n    key: &quot;validate_shape&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/kernel/read&quot;\\n  op: &quot;Identity&quot;\\n  input: &quot;conv2d/kernel&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d/kernel&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/bias/Initializer/zeros&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d/bias&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_FLOAT\\n        tensor_shape {\\n          dim {\\n            size: 6\\n          }\\n        }\\n        float_val: 0.0\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/bias&quot;\\n  op: &quot;VariableV2&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d/bias&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;container&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;shape&quot;\\n    value {\\n      shape {\\n        dim {\\n          size: 6\\n        }\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;shared_name&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/bias/Assign&quot;\\n  op: &quot;Assign&quot;\\n  input: &quot;conv2d/bias&quot;\\n  input: &quot;conv2d/bias/Initializer/zeros&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d/bias&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;use_locking&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n  attr {\\n    key: &quot;validate_shape&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/bias/read&quot;\\n  op: &quot;Identity&quot;\\n  input: &quot;conv2d/bias&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d/bias&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/dilation_rate&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_INT32\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_INT32\\n        tensor_shape {\\n          dim {\\n            size: 2\\n          }\\n        }\\n        tensor_content: &quot;\\\\001\\\\000\\\\000\\\\000\\\\001\\\\000\\\\000\\\\000&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/Conv2D&quot;\\n  op: &quot;Conv2D&quot;\\n  input: &quot;input&quot;\\n  input: &quot;conv2d/kernel/read&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;data_format&quot;\\n    value {\\n      s: &quot;NHWC&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;dilations&quot;\\n    value {\\n      list {\\n        i: 1\\n        i: 1\\n        i: 1\\n        i: 1\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;padding&quot;\\n    value {\\n      s: &quot;VALID&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;strides&quot;\\n    value {\\n      list {\\n        i: 1\\n        i: 1\\n        i: 1\\n        i: 1\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;use_cudnn_on_gpu&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d/BiasAdd&quot;\\n  op: &quot;BiasAdd&quot;\\n  input: &quot;conv2d/Conv2D&quot;\\n  input: &quot;conv2d/bias/read&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;data_format&quot;\\n    value {\\n      s: &quot;NHWC&quot;\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/kernel/Initializer/random_uniform/shape&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d_1/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_INT32\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_INT32\\n        tensor_shape {\\n          dim {\\n            size: 4\\n          }\\n        }\\n        tensor_content: &quot;\\\\003\\\\000\\\\000\\\\000\\\\003\\\\000\\\\000\\\\000\\\\006\\\\000\\\\000\\\\000\\\\006\\\\000\\\\000\\\\000&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/kernel/Initializer/random_uniform/min&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d_1/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_FLOAT\\n        tensor_shape {\\n        }\\n        float_val: -0.2357022613286972\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/kernel/Initializer/random_uniform/max&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d_1/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_FLOAT\\n        tensor_shape {\\n        }\\n        float_val: 0.2357022613286972\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/kernel/Initializer/random_uniform/RandomUniform&quot;\\n  op: &quot;RandomUniform&quot;\\n  input: &quot;conv2d_1/kernel/Initializer/random_uniform/shape&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_INT32\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d_1/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;seed&quot;\\n    value {\\n      i: 0\\n    }\\n  }\\n  attr {\\n    key: &quot;seed2&quot;\\n    value {\\n      i: 0\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/kernel/Initializer/random_uniform/sub&quot;\\n  op: &quot;Sub&quot;\\n  input: &quot;conv2d_1/kernel/Initializer/random_uniform/max&quot;\\n  input: &quot;conv2d_1/kernel/Initializer/random_uniform/min&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d_1/kernel&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/kernel/Initializer/random_uniform/mul&quot;\\n  op: &quot;Mul&quot;\\n  input: &quot;conv2d_1/kernel/Initializer/random_uniform/RandomUniform&quot;\\n  input: &quot;conv2d_1/kernel/Initializer/random_uniform/sub&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d_1/kernel&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/kernel/Initializer/random_uniform&quot;\\n  op: &quot;Add&quot;\\n  input: &quot;conv2d_1/kernel/Initializer/random_uniform/mul&quot;\\n  input: &quot;conv2d_1/kernel/Initializer/random_uniform/min&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d_1/kernel&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/kernel&quot;\\n  op: &quot;VariableV2&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d_1/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;container&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;shape&quot;\\n    value {\\n      shape {\\n        dim {\\n          size: 3\\n        }\\n        dim {\\n          size: 3\\n        }\\n        dim {\\n          size: 6\\n        }\\n        dim {\\n          size: 6\\n        }\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;shared_name&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/kernel/Assign&quot;\\n  op: &quot;Assign&quot;\\n  input: &quot;conv2d_1/kernel&quot;\\n  input: &quot;conv2d_1/kernel/Initializer/random_uniform&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d_1/kernel&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;use_locking&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n  attr {\\n    key: &quot;validate_shape&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/kernel/read&quot;\\n  op: &quot;Identity&quot;\\n  input: &quot;conv2d_1/kernel&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d_1/kernel&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/bias/Initializer/zeros&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d_1/bias&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_FLOAT\\n        tensor_shape {\\n          dim {\\n            size: 6\\n          }\\n        }\\n        float_val: 0.0\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/bias&quot;\\n  op: &quot;VariableV2&quot;\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d_1/bias&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;container&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;shape&quot;\\n    value {\\n      shape {\\n        dim {\\n          size: 6\\n        }\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;shared_name&quot;\\n    value {\\n      s: &quot;&quot;\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/bias/Assign&quot;\\n  op: &quot;Assign&quot;\\n  input: &quot;conv2d_1/bias&quot;\\n  input: &quot;conv2d_1/bias/Initializer/zeros&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d_1/bias&quot;\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;use_locking&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n  attr {\\n    key: &quot;validate_shape&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/bias/read&quot;\\n  op: &quot;Identity&quot;\\n  input: &quot;conv2d_1/bias&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;_class&quot;\\n    value {\\n      list {\\n        s: &quot;loc:@conv2d_1/bias&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/dilation_rate&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_INT32\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_INT32\\n        tensor_shape {\\n          dim {\\n            size: 2\\n          }\\n        }\\n        tensor_content: &quot;\\\\001\\\\000\\\\000\\\\000\\\\001\\\\000\\\\000\\\\000&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/Conv2D&quot;\\n  op: &quot;Conv2D&quot;\\n  input: &quot;conv2d/BiasAdd&quot;\\n  input: &quot;conv2d_1/kernel/read&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;data_format&quot;\\n    value {\\n      s: &quot;NHWC&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;dilations&quot;\\n    value {\\n      list {\\n        i: 1\\n        i: 1\\n        i: 1\\n        i: 1\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;padding&quot;\\n    value {\\n      s: &quot;VALID&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;strides&quot;\\n    value {\\n      list {\\n        i: 1\\n        i: 1\\n        i: 1\\n        i: 1\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;use_cudnn_on_gpu&quot;\\n    value {\\n      b: true\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;conv2d_1/BiasAdd&quot;\\n  op: &quot;BiasAdd&quot;\\n  input: &quot;conv2d_1/Conv2D&quot;\\n  input: &quot;conv2d_1/bias/read&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;data_format&quot;\\n    value {\\n      s: &quot;NHWC&quot;\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;max_pooling2d/MaxPool&quot;\\n  op: &quot;MaxPool&quot;\\n  input: &quot;conv2d_1/BiasAdd&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;data_format&quot;\\n    value {\\n      s: &quot;NHWC&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;ksize&quot;\\n    value {\\n      list {\\n        i: 1\\n        i: 1\\n        i: 1\\n        i: 1\\n      }\\n    }\\n  }\\n  attr {\\n    key: &quot;padding&quot;\\n    value {\\n      s: &quot;VALID&quot;\\n    }\\n  }\\n  attr {\\n    key: &quot;strides&quot;\\n    value {\\n      list {\\n        i: 1\\n        i: 2\\n        i: 2\\n        i: 1\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;output/shape&quot;\\n  op: &quot;Const&quot;\\n  attr {\\n    key: &quot;dtype&quot;\\n    value {\\n      type: DT_INT32\\n    }\\n  }\\n  attr {\\n    key: &quot;value&quot;\\n    value {\\n      tensor {\\n        dtype: DT_INT32\\n        tensor_shape {\\n          dim {\\n            size: 4\\n          }\\n        }\\n        tensor_content: &quot;\\\\001\\\\000\\\\000\\\\000n\\\\000\\\\000\\\\000n\\\\000\\\\000\\\\000\\\\006\\\\000\\\\000\\\\000&quot;\\n      }\\n    }\\n  }\\n}\\nnode {\\n  name: &quot;output&quot;\\n  op: &quot;Reshape&quot;\\n  input: &quot;max_pooling2d/MaxPool&quot;\\n  input: &quot;output/shape&quot;\\n  attr {\\n    key: &quot;T&quot;\\n    value {\\n      type: DT_FLOAT\\n    }\\n  }\\n  attr {\\n    key: &quot;Tshape&quot;\\n    value {\\n      type: DT_INT32\\n    }\\n  }\\n}\\n';\n",
       "          }\n",
       "        </script>\n",
       "        <link rel=&quot;import&quot; href=&quot;https://tensorboard.appspot.com/tf-graph-basic.build.html&quot; onload=load()>\n",
       "        <div style=&quot;height:600px&quot;>\n",
       "          <tf-graph-basic id=&quot;graph0.13782732954427268&quot;></tf-graph-basic>\n",
       "        </div>\n",
       "    \"></iframe>\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "tf.reset_default_graph()\n",
    "input_image = tf.placeholder(tf.float32, shape=(None,224,224,3), name='input')\n",
    "conv1 = tf.layers.conv2d(input_image, 6, (3,3)) # first 6 convolutions\n",
    "conv2 = tf.layers.conv2d(conv1, 6, (3,3))       # second 6 convolutions\n",
    "\n",
    "# a max pool (1, 1) with (2, 2) strides\n",
    "maxpool = tf.layers.max_pooling2d(conv2, (1,1), (2,2))\n",
    "\n",
    "# define an output, the shape must correspond to the\n",
    "# number of values that are output from conv2 = 6\n",
    "output = tf.reshape(maxpool, [1, 110, 110, 6], name='output')\n",
    "\n",
    "with tf.Session() as sess:\n",
    "    show_graph(sess.graph.as_graph_def())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can make a test, one more time we open an image, send it to \"input\" and we will read the \"output\" operation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Open an image, resize it to (224, 224) that is the shape of the graph input\n",
    "image = PIL.Image.open('data/train/burger/02821fe8e4e7bf47046e1249e840d9b5.jpeg')\n",
    "image = image.resize((224,224))\n",
    "\n",
    "# use numpy to reshape the array and to add a dimension\n",
    "image = np.array(image)\n",
    "image = np.expand_dims(image, axis=0)\n",
    "\n",
    "# open a session, initialize variables and inject the image\n",
    "with tf.Session() as sess:\n",
    "    init = tf.global_variables_initializer()\n",
    "    sess.run(init)\n",
    "    res = sess.run('output:0', {\n",
    "        'input:0' : image\n",
    "    })\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "res = np.squeeze(res)\n",
    "\n",
    "columns = 4\n",
    "rows = math.ceil(res.shape[2] / columns)\n",
    "\n",
    "plt.figure(figsize=(12, 6))\n",
    "for i in range(res.shape[2]):\n",
    "    plt.subplot(\"%d%d%d\" % (rows, columns, i+1))\n",
    "    plt.imshow(res[:,:,i])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is what does Tensorflow. We inject variables, placeholders, operations... and we can \"flow\" data inside tensors and operation in a Session."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}