python_tinyML/02-first_assigment.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "01-Assigment.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyOnuQU7xR+Z6L9nmKsTcIZT",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/ferris77/my_notebooks/blob/main/python_tinyML/02-first_assigment.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uOvTcb7BZb9j"
      },
      "source": [
        "# We first import TensorFlow and other libraries\n",
        "import tensorflow as tf\n",
        "import numpy as np\n",
        "from tensorflow.keras import Sequential\n",
        "from tensorflow.keras.layers import Dense\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# We then set up some functions and local variables\n",
        "predictions = []\n",
        "class myCallback(tf.keras.callbacks.Callback):\n",
        "  def on_epoch_end(self, epoch, logs={}):\n",
        "    predictions.append(model.predict(xs))\n",
        "callbacks = myCallback()\n",
        "\n",
        "# We then define the xs (inputs) and ys (outputs)\n",
        "xs = np.array([-1.0, 0.0, 1.0, 2.0, 3.0, 4.0], dtype=float)\n",
        "ys = np.array([-3.0, -1.0, 1.0, 3.0, 5.0, 7.0], dtype=float)"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zIY0TVNuZjrC"
      },
      "source": [
        "SHAPE = [1] #YOUR CODE HERE#\n",
        "LOSS = 'mean_squared_error' #YOUR CODE HERE#\n",
        "\n",
        "# Define your model type\n",
        "model = Sequential([Dense(units=1, input_shape=SHAPE)])\n",
        "    \n",
        "# Compile your model with choice of optimizer and loss function\n",
        "model.compile(optimizer='sgd', loss=LOSS)"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XY2qseZRZltV",
        "outputId": "c49704a4-3933-4f14-890a-53e8c1654283"
      },
      "source": [
        "# We then fit the model\n",
        "model.fit(xs, ys, epochs=300, callbacks=[callbacks], verbose=2)"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/300\n",
            "1/1 - 0s - loss: 0.5878\n",
            "Epoch 2/300\n",
            "1/1 - 0s - loss: 0.5746\n",
            "Epoch 3/300\n",
            "1/1 - 0s - loss: 0.5618\n",
            "Epoch 4/300\n",
            "1/1 - 0s - loss: 0.5495\n",
            "Epoch 5/300\n",
            "1/1 - 0s - loss: 0.5376\n",
            "Epoch 6/300\n",
            "1/1 - 0s - loss: 0.5261\n",
            "Epoch 7/300\n",
            "1/1 - 0s - loss: 0.5150\n",
            "Epoch 8/300\n",
            "1/1 - 0s - loss: 0.5041\n",
            "Epoch 9/300\n",
            "1/1 - 0s - loss: 0.4935\n",
            "Epoch 10/300\n",
            "1/1 - 0s - loss: 0.4832\n",
            "Epoch 11/300\n",
            "1/1 - 0s - loss: 0.4732\n",
            "Epoch 12/300\n",
            "1/1 - 0s - loss: 0.4633\n",
            "Epoch 13/300\n",
            "1/1 - 0s - loss: 0.4537\n",
            "Epoch 14/300\n",
            "1/1 - 0s - loss: 0.4443\n",
            "Epoch 15/300\n",
            "1/1 - 0s - loss: 0.4352\n",
            "Epoch 16/300\n",
            "1/1 - 0s - loss: 0.4262\n",
            "Epoch 17/300\n",
            "1/1 - 0s - loss: 0.4174\n",
            "Epoch 18/300\n",
            "1/1 - 0s - loss: 0.4088\n",
            "Epoch 19/300\n",
            "1/1 - 0s - loss: 0.4004\n",
            "Epoch 20/300\n",
            "1/1 - 0s - loss: 0.3921\n",
            "Epoch 21/300\n",
            "1/1 - 0s - loss: 0.3841\n",
            "Epoch 22/300\n",
            "1/1 - 0s - loss: 0.3762\n",
            "Epoch 23/300\n",
            "1/1 - 0s - loss: 0.3684\n",
            "Epoch 24/300\n",
            "1/1 - 0s - loss: 0.3609\n",
            "Epoch 25/300\n",
            "1/1 - 0s - loss: 0.3534\n",
            "Epoch 26/300\n",
            "1/1 - 0s - loss: 0.3462\n",
            "Epoch 27/300\n",
            "1/1 - 0s - loss: 0.3391\n",
            "Epoch 28/300\n",
            "1/1 - 0s - loss: 0.3321\n",
            "Epoch 29/300\n",
            "1/1 - 0s - loss: 0.3253\n",
            "Epoch 30/300\n",
            "1/1 - 0s - loss: 0.3186\n",
            "Epoch 31/300\n",
            "1/1 - 0s - loss: 0.3120\n",
            "Epoch 32/300\n",
            "1/1 - 0s - loss: 0.3056\n",
            "Epoch 33/300\n",
            "1/1 - 0s - loss: 0.2994\n",
            "Epoch 34/300\n",
            "1/1 - 0s - loss: 0.2932\n",
            "Epoch 35/300\n",
            "1/1 - 0s - loss: 0.2872\n",
            "Epoch 36/300\n",
            "1/1 - 0s - loss: 0.2813\n",
            "Epoch 37/300\n",
            "1/1 - 0s - loss: 0.2755\n",
            "Epoch 38/300\n",
            "1/1 - 0s - loss: 0.2698\n",
            "Epoch 39/300\n",
            "1/1 - 0s - loss: 0.2643\n",
            "Epoch 40/300\n",
            "1/1 - 0s - loss: 0.2589\n",
            "Epoch 41/300\n",
            "1/1 - 0s - loss: 0.2536\n",
            "Epoch 42/300\n",
            "1/1 - 0s - loss: 0.2483\n",
            "Epoch 43/300\n",
            "1/1 - 0s - loss: 0.2432\n",
            "Epoch 44/300\n",
            "1/1 - 0s - loss: 0.2382\n",
            "Epoch 45/300\n",
            "1/1 - 0s - loss: 0.2334\n",
            "Epoch 46/300\n",
            "1/1 - 0s - loss: 0.2286\n",
            "Epoch 47/300\n",
            "1/1 - 0s - loss: 0.2239\n",
            "Epoch 48/300\n",
            "1/1 - 0s - loss: 0.2193\n",
            "Epoch 49/300\n",
            "1/1 - 0s - loss: 0.2148\n",
            "Epoch 50/300\n",
            "1/1 - 0s - loss: 0.2104\n",
            "Epoch 51/300\n",
            "1/1 - 0s - loss: 0.2060\n",
            "Epoch 52/300\n",
            "1/1 - 0s - loss: 0.2018\n",
            "Epoch 53/300\n",
            "1/1 - 0s - loss: 0.1977\n",
            "Epoch 54/300\n",
            "1/1 - 0s - loss: 0.1936\n",
            "Epoch 55/300\n",
            "1/1 - 0s - loss: 0.1896\n",
            "Epoch 56/300\n",
            "1/1 - 0s - loss: 0.1857\n",
            "Epoch 57/300\n",
            "1/1 - 0s - loss: 0.1819\n",
            "Epoch 58/300\n",
            "1/1 - 0s - loss: 0.1782\n",
            "Epoch 59/300\n",
            "1/1 - 0s - loss: 0.1745\n",
            "Epoch 60/300\n",
            "1/1 - 0s - loss: 0.1709\n",
            "Epoch 61/300\n",
            "1/1 - 0s - loss: 0.1674\n",
            "Epoch 62/300\n",
            "1/1 - 0s - loss: 0.1640\n",
            "Epoch 63/300\n",
            "1/1 - 0s - loss: 0.1606\n",
            "Epoch 64/300\n",
            "1/1 - 0s - loss: 0.1573\n",
            "Epoch 65/300\n",
            "1/1 - 0s - loss: 0.1541\n",
            "Epoch 66/300\n",
            "1/1 - 0s - loss: 0.1509\n",
            "Epoch 67/300\n",
            "1/1 - 0s - loss: 0.1478\n",
            "Epoch 68/300\n",
            "1/1 - 0s - loss: 0.1448\n",
            "Epoch 69/300\n",
            "1/1 - 0s - loss: 0.1418\n",
            "Epoch 70/300\n",
            "1/1 - 0s - loss: 0.1389\n",
            "Epoch 71/300\n",
            "1/1 - 0s - loss: 0.1360\n",
            "Epoch 72/300\n",
            "1/1 - 0s - loss: 0.1332\n",
            "Epoch 73/300\n",
            "1/1 - 0s - loss: 0.1305\n",
            "Epoch 74/300\n",
            "1/1 - 0s - loss: 0.1278\n",
            "Epoch 75/300\n",
            "1/1 - 0s - loss: 0.1252\n",
            "Epoch 76/300\n",
            "1/1 - 0s - loss: 0.1226\n",
            "Epoch 77/300\n",
            "1/1 - 0s - loss: 0.1201\n",
            "Epoch 78/300\n",
            "1/1 - 0s - loss: 0.1176\n",
            "Epoch 79/300\n",
            "1/1 - 0s - loss: 0.1152\n",
            "Epoch 80/300\n",
            "1/1 - 0s - loss: 0.1129\n",
            "Epoch 81/300\n",
            "1/1 - 0s - loss: 0.1105\n",
            "Epoch 82/300\n",
            "1/1 - 0s - loss: 0.1083\n",
            "Epoch 83/300\n",
            "1/1 - 0s - loss: 0.1060\n",
            "Epoch 84/300\n",
            "1/1 - 0s - loss: 0.1039\n",
            "Epoch 85/300\n",
            "1/1 - 0s - loss: 0.1017\n",
            "Epoch 86/300\n",
            "1/1 - 0s - loss: 0.0996\n",
            "Epoch 87/300\n",
            "1/1 - 0s - loss: 0.0976\n",
            "Epoch 88/300\n",
            "1/1 - 0s - loss: 0.0956\n",
            "Epoch 89/300\n",
            "1/1 - 0s - loss: 0.0936\n",
            "Epoch 90/300\n",
            "1/1 - 0s - loss: 0.0917\n",
            "Epoch 91/300\n",
            "1/1 - 0s - loss: 0.0898\n",
            "Epoch 92/300\n",
            "1/1 - 0s - loss: 0.0880\n",
            "Epoch 93/300\n",
            "1/1 - 0s - loss: 0.0862\n",
            "Epoch 94/300\n",
            "1/1 - 0s - loss: 0.0844\n",
            "Epoch 95/300\n",
            "1/1 - 0s - loss: 0.0827\n",
            "Epoch 96/300\n",
            "1/1 - 0s - loss: 0.0810\n",
            "Epoch 97/300\n",
            "1/1 - 0s - loss: 0.0793\n",
            "Epoch 98/300\n",
            "1/1 - 0s - loss: 0.0777\n",
            "Epoch 99/300\n",
            "1/1 - 0s - loss: 0.0761\n",
            "Epoch 100/300\n",
            "1/1 - 0s - loss: 0.0745\n",
            "Epoch 101/300\n",
            "1/1 - 0s - loss: 0.0730\n",
            "Epoch 102/300\n",
            "1/1 - 0s - loss: 0.0715\n",
            "Epoch 103/300\n",
            "1/1 - 0s - loss: 0.0700\n",
            "Epoch 104/300\n",
            "1/1 - 0s - loss: 0.0686\n",
            "Epoch 105/300\n",
            "1/1 - 0s - loss: 0.0672\n",
            "Epoch 106/300\n",
            "1/1 - 0s - loss: 0.0658\n",
            "Epoch 107/300\n",
            "1/1 - 0s - loss: 0.0644\n",
            "Epoch 108/300\n",
            "1/1 - 0s - loss: 0.0631\n",
            "Epoch 109/300\n",
            "1/1 - 0s - loss: 0.0618\n",
            "Epoch 110/300\n",
            "1/1 - 0s - loss: 0.0606\n",
            "Epoch 111/300\n",
            "1/1 - 0s - loss: 0.0593\n",
            "Epoch 112/300\n",
            "1/1 - 0s - loss: 0.0581\n",
            "Epoch 113/300\n",
            "1/1 - 0s - loss: 0.0569\n",
            "Epoch 114/300\n",
            "1/1 - 0s - loss: 0.0557\n",
            "Epoch 115/300\n",
            "1/1 - 0s - loss: 0.0546\n",
            "Epoch 116/300\n",
            "1/1 - 0s - loss: 0.0535\n",
            "Epoch 117/300\n",
            "1/1 - 0s - loss: 0.0524\n",
            "Epoch 118/300\n",
            "1/1 - 0s - loss: 0.0513\n",
            "Epoch 119/300\n",
            "1/1 - 0s - loss: 0.0502\n",
            "Epoch 120/300\n",
            "1/1 - 0s - loss: 0.0492\n",
            "Epoch 121/300\n",
            "1/1 - 0s - loss: 0.0482\n",
            "Epoch 122/300\n",
            "1/1 - 0s - loss: 0.0472\n",
            "Epoch 123/300\n",
            "1/1 - 0s - loss: 0.0462\n",
            "Epoch 124/300\n",
            "1/1 - 0s - loss: 0.0453\n",
            "Epoch 125/300\n",
            "1/1 - 0s - loss: 0.0444\n",
            "Epoch 126/300\n",
            "1/1 - 0s - loss: 0.0434\n",
            "Epoch 127/300\n",
            "1/1 - 0s - loss: 0.0426\n",
            "Epoch 128/300\n",
            "1/1 - 0s - loss: 0.0417\n",
            "Epoch 129/300\n",
            "1/1 - 0s - loss: 0.0408\n",
            "Epoch 130/300\n",
            "1/1 - 0s - loss: 0.0400\n",
            "Epoch 131/300\n",
            "1/1 - 0s - loss: 0.0392\n",
            "Epoch 132/300\n",
            "1/1 - 0s - loss: 0.0384\n",
            "Epoch 133/300\n",
            "1/1 - 0s - loss: 0.0376\n",
            "Epoch 134/300\n",
            "1/1 - 0s - loss: 0.0368\n",
            "Epoch 135/300\n",
            "1/1 - 0s - loss: 0.0360\n",
            "Epoch 136/300\n",
            "1/1 - 0s - loss: 0.0353\n",
            "Epoch 137/300\n",
            "1/1 - 0s - loss: 0.0346\n",
            "Epoch 138/300\n",
            "1/1 - 0s - loss: 0.0339\n",
            "Epoch 139/300\n",
            "1/1 - 0s - loss: 0.0332\n",
            "Epoch 140/300\n",
            "1/1 - 0s - loss: 0.0325\n",
            "Epoch 141/300\n",
            "1/1 - 0s - loss: 0.0318\n",
            "Epoch 142/300\n",
            "1/1 - 0s - loss: 0.0312\n",
            "Epoch 143/300\n",
            "1/1 - 0s - loss: 0.0305\n",
            "Epoch 144/300\n",
            "1/1 - 0s - loss: 0.0299\n",
            "Epoch 145/300\n",
            "1/1 - 0s - loss: 0.0293\n",
            "Epoch 146/300\n",
            "1/1 - 0s - loss: 0.0287\n",
            "Epoch 147/300\n",
            "1/1 - 0s - loss: 0.0281\n",
            "Epoch 148/300\n",
            "1/1 - 0s - loss: 0.0275\n",
            "Epoch 149/300\n",
            "1/1 - 0s - loss: 0.0270\n",
            "Epoch 150/300\n",
            "1/1 - 0s - loss: 0.0264\n",
            "Epoch 151/300\n",
            "1/1 - 0s - loss: 0.0259\n",
            "Epoch 152/300\n",
            "1/1 - 0s - loss: 0.0253\n",
            "Epoch 153/300\n",
            "1/1 - 0s - loss: 0.0248\n",
            "Epoch 154/300\n",
            "1/1 - 0s - loss: 0.0243\n",
            "Epoch 155/300\n",
            "1/1 - 0s - loss: 0.0238\n",
            "Epoch 156/300\n",
            "1/1 - 0s - loss: 0.0233\n",
            "Epoch 157/300\n",
            "1/1 - 0s - loss: 0.0228\n",
            "Epoch 158/300\n",
            "1/1 - 0s - loss: 0.0224\n",
            "Epoch 159/300\n",
            "1/1 - 0s - loss: 0.0219\n",
            "Epoch 160/300\n",
            "1/1 - 0s - loss: 0.0215\n",
            "Epoch 161/300\n",
            "1/1 - 0s - loss: 0.0210\n",
            "Epoch 162/300\n",
            "1/1 - 0s - loss: 0.0206\n",
            "Epoch 163/300\n",
            "1/1 - 0s - loss: 0.0202\n",
            "Epoch 164/300\n",
            "1/1 - 0s - loss: 0.0197\n",
            "Epoch 165/300\n",
            "1/1 - 0s - loss: 0.0193\n",
            "Epoch 166/300\n",
            "1/1 - 0s - loss: 0.0189\n",
            "Epoch 167/300\n",
            "1/1 - 0s - loss: 0.0186\n",
            "Epoch 168/300\n",
            "1/1 - 0s - loss: 0.0182\n",
            "Epoch 169/300\n",
            "1/1 - 0s - loss: 0.0178\n",
            "Epoch 170/300\n",
            "1/1 - 0s - loss: 0.0174\n",
            "Epoch 171/300\n",
            "1/1 - 0s - loss: 0.0171\n",
            "Epoch 172/300\n",
            "1/1 - 0s - loss: 0.0167\n",
            "Epoch 173/300\n",
            "1/1 - 0s - loss: 0.0164\n",
            "Epoch 174/300\n",
            "1/1 - 0s - loss: 0.0160\n",
            "Epoch 175/300\n",
            "1/1 - 0s - loss: 0.0157\n",
            "Epoch 176/300\n",
            "1/1 - 0s - loss: 0.0154\n",
            "Epoch 177/300\n",
            "1/1 - 0s - loss: 0.0151\n",
            "Epoch 178/300\n",
            "1/1 - 0s - loss: 0.0148\n",
            "Epoch 179/300\n",
            "1/1 - 0s - loss: 0.0145\n",
            "Epoch 180/300\n",
            "1/1 - 0s - loss: 0.0142\n",
            "Epoch 181/300\n",
            "1/1 - 0s - loss: 0.0139\n",
            "Epoch 182/300\n",
            "1/1 - 0s - loss: 0.0136\n",
            "Epoch 183/300\n",
            "1/1 - 0s - loss: 0.0133\n",
            "Epoch 184/300\n",
            "1/1 - 0s - loss: 0.0130\n",
            "Epoch 185/300\n",
            "1/1 - 0s - loss: 0.0128\n",
            "Epoch 186/300\n",
            "1/1 - 0s - loss: 0.0125\n",
            "Epoch 187/300\n",
            "1/1 - 0s - loss: 0.0122\n",
            "Epoch 188/300\n",
            "1/1 - 0s - loss: 0.0120\n",
            "Epoch 189/300\n",
            "1/1 - 0s - loss: 0.0118\n",
            "Epoch 190/300\n",
            "1/1 - 0s - loss: 0.0115\n",
            "Epoch 191/300\n",
            "1/1 - 0s - loss: 0.0113\n",
            "Epoch 192/300\n",
            "1/1 - 0s - loss: 0.0110\n",
            "Epoch 193/300\n",
            "1/1 - 0s - loss: 0.0108\n",
            "Epoch 194/300\n",
            "1/1 - 0s - loss: 0.0106\n",
            "Epoch 195/300\n",
            "1/1 - 0s - loss: 0.0104\n",
            "Epoch 196/300\n",
            "1/1 - 0s - loss: 0.0102\n",
            "Epoch 197/300\n",
            "1/1 - 0s - loss: 0.0100\n",
            "Epoch 198/300\n",
            "1/1 - 0s - loss: 0.0097\n",
            "Epoch 199/300\n",
            "1/1 - 0s - loss: 0.0095\n",
            "Epoch 200/300\n",
            "1/1 - 0s - loss: 0.0094\n",
            "Epoch 201/300\n",
            "1/1 - 0s - loss: 0.0092\n",
            "Epoch 202/300\n",
            "1/1 - 0s - loss: 0.0090\n",
            "Epoch 203/300\n",
            "1/1 - 0s - loss: 0.0088\n",
            "Epoch 204/300\n",
            "1/1 - 0s - loss: 0.0086\n",
            "Epoch 205/300\n",
            "1/1 - 0s - loss: 0.0084\n",
            "Epoch 206/300\n",
            "1/1 - 0s - loss: 0.0083\n",
            "Epoch 207/300\n",
            "1/1 - 0s - loss: 0.0081\n",
            "Epoch 208/300\n",
            "1/1 - 0s - loss: 0.0079\n",
            "Epoch 209/300\n",
            "1/1 - 0s - loss: 0.0078\n",
            "Epoch 210/300\n",
            "1/1 - 0s - loss: 0.0076\n",
            "Epoch 211/300\n",
            "1/1 - 0s - loss: 0.0074\n",
            "Epoch 212/300\n",
            "1/1 - 0s - loss: 0.0073\n",
            "Epoch 213/300\n",
            "1/1 - 0s - loss: 0.0071\n",
            "Epoch 214/300\n",
            "1/1 - 0s - loss: 0.0070\n",
            "Epoch 215/300\n",
            "1/1 - 0s - loss: 0.0069\n",
            "Epoch 216/300\n",
            "1/1 - 0s - loss: 0.0067\n",
            "Epoch 217/300\n",
            "1/1 - 0s - loss: 0.0066\n",
            "Epoch 218/300\n",
            "1/1 - 0s - loss: 0.0064\n",
            "Epoch 219/300\n",
            "1/1 - 0s - loss: 0.0063\n",
            "Epoch 220/300\n",
            "1/1 - 0s - loss: 0.0062\n",
            "Epoch 221/300\n",
            "1/1 - 0s - loss: 0.0060\n",
            "Epoch 222/300\n",
            "1/1 - 0s - loss: 0.0059\n",
            "Epoch 223/300\n",
            "1/1 - 0s - loss: 0.0058\n",
            "Epoch 224/300\n",
            "1/1 - 0s - loss: 0.0057\n",
            "Epoch 225/300\n",
            "1/1 - 0s - loss: 0.0056\n",
            "Epoch 226/300\n",
            "1/1 - 0s - loss: 0.0055\n",
            "Epoch 227/300\n",
            "1/1 - 0s - loss: 0.0053\n",
            "Epoch 228/300\n",
            "1/1 - 0s - loss: 0.0052\n",
            "Epoch 229/300\n",
            "1/1 - 0s - loss: 0.0051\n",
            "Epoch 230/300\n",
            "1/1 - 0s - loss: 0.0050\n",
            "Epoch 231/300\n",
            "1/1 - 0s - loss: 0.0049\n",
            "Epoch 232/300\n",
            "1/1 - 0s - loss: 0.0048\n",
            "Epoch 233/300\n",
            "1/1 - 0s - loss: 0.0047\n",
            "Epoch 234/300\n",
            "1/1 - 0s - loss: 0.0046\n",
            "Epoch 235/300\n",
            "1/1 - 0s - loss: 0.0045\n",
            "Epoch 236/300\n",
            "1/1 - 0s - loss: 0.0044\n",
            "Epoch 237/300\n",
            "1/1 - 0s - loss: 0.0043\n",
            "Epoch 238/300\n",
            "1/1 - 0s - loss: 0.0043\n",
            "Epoch 239/300\n",
            "1/1 - 0s - loss: 0.0042\n",
            "Epoch 240/300\n",
            "1/1 - 0s - loss: 0.0041\n",
            "Epoch 241/300\n",
            "1/1 - 0s - loss: 0.0040\n",
            "Epoch 242/300\n",
            "1/1 - 0s - loss: 0.0039\n",
            "Epoch 243/300\n",
            "1/1 - 0s - loss: 0.0038\n",
            "Epoch 244/300\n",
            "1/1 - 0s - loss: 0.0038\n",
            "Epoch 245/300\n",
            "1/1 - 0s - loss: 0.0037\n",
            "Epoch 246/300\n",
            "1/1 - 0s - loss: 0.0036\n",
            "Epoch 247/300\n",
            "1/1 - 0s - loss: 0.0035\n",
            "Epoch 248/300\n",
            "1/1 - 0s - loss: 0.0035\n",
            "Epoch 249/300\n",
            "1/1 - 0s - loss: 0.0034\n",
            "Epoch 250/300\n",
            "1/1 - 0s - loss: 0.0033\n",
            "Epoch 251/300\n",
            "1/1 - 0s - loss: 0.0032\n",
            "Epoch 252/300\n",
            "1/1 - 0s - loss: 0.0032\n",
            "Epoch 253/300\n",
            "1/1 - 0s - loss: 0.0031\n",
            "Epoch 254/300\n",
            "1/1 - 0s - loss: 0.0030\n",
            "Epoch 255/300\n",
            "1/1 - 0s - loss: 0.0030\n",
            "Epoch 256/300\n",
            "1/1 - 0s - loss: 0.0029\n",
            "Epoch 257/300\n",
            "1/1 - 0s - loss: 0.0029\n",
            "Epoch 258/300\n",
            "1/1 - 0s - loss: 0.0028\n",
            "Epoch 259/300\n",
            "1/1 - 0s - loss: 0.0027\n",
            "Epoch 260/300\n",
            "1/1 - 0s - loss: 0.0027\n",
            "Epoch 261/300\n",
            "1/1 - 0s - loss: 0.0026\n",
            "Epoch 262/300\n",
            "1/1 - 0s - loss: 0.0026\n",
            "Epoch 263/300\n",
            "1/1 - 0s - loss: 0.0025\n",
            "Epoch 264/300\n",
            "1/1 - 0s - loss: 0.0025\n",
            "Epoch 265/300\n",
            "1/1 - 0s - loss: 0.0024\n",
            "Epoch 266/300\n",
            "1/1 - 0s - loss: 0.0024\n",
            "Epoch 267/300\n",
            "1/1 - 0s - loss: 0.0023\n",
            "Epoch 268/300\n",
            "1/1 - 0s - loss: 0.0023\n",
            "Epoch 269/300\n",
            "1/1 - 0s - loss: 0.0022\n",
            "Epoch 270/300\n",
            "1/1 - 0s - loss: 0.0022\n",
            "Epoch 271/300\n",
            "1/1 - 0s - loss: 0.0021\n",
            "Epoch 272/300\n",
            "1/1 - 0s - loss: 0.0021\n",
            "Epoch 273/300\n",
            "1/1 - 0s - loss: 0.0021\n",
            "Epoch 274/300\n",
            "1/1 - 0s - loss: 0.0020\n",
            "Epoch 275/300\n",
            "1/1 - 0s - loss: 0.0020\n",
            "Epoch 276/300\n",
            "1/1 - 0s - loss: 0.0019\n",
            "Epoch 277/300\n",
            "1/1 - 0s - loss: 0.0019\n",
            "Epoch 278/300\n",
            "1/1 - 0s - loss: 0.0019\n",
            "Epoch 279/300\n",
            "1/1 - 0s - loss: 0.0018\n",
            "Epoch 280/300\n",
            "1/1 - 0s - loss: 0.0018\n",
            "Epoch 281/300\n",
            "1/1 - 0s - loss: 0.0017\n",
            "Epoch 282/300\n",
            "1/1 - 0s - loss: 0.0017\n",
            "Epoch 283/300\n",
            "1/1 - 0s - loss: 0.0017\n",
            "Epoch 284/300\n",
            "1/1 - 0s - loss: 0.0016\n",
            "Epoch 285/300\n",
            "1/1 - 0s - loss: 0.0016\n",
            "Epoch 286/300\n",
            "1/1 - 0s - loss: 0.0016\n",
            "Epoch 287/300\n",
            "1/1 - 0s - loss: 0.0015\n",
            "Epoch 288/300\n",
            "1/1 - 0s - loss: 0.0015\n",
            "Epoch 289/300\n",
            "1/1 - 0s - loss: 0.0015\n",
            "Epoch 290/300\n",
            "1/1 - 0s - loss: 0.0014\n",
            "Epoch 291/300\n",
            "1/1 - 0s - loss: 0.0014\n",
            "Epoch 292/300\n",
            "1/1 - 0s - loss: 0.0014\n",
            "Epoch 293/300\n",
            "1/1 - 0s - loss: 0.0014\n",
            "Epoch 294/300\n",
            "1/1 - 0s - loss: 0.0013\n",
            "Epoch 295/300\n",
            "1/1 - 0s - loss: 0.0013\n",
            "Epoch 296/300\n",
            "1/1 - 0s - loss: 0.0013\n",
            "Epoch 297/300\n",
            "1/1 - 0s - loss: 0.0012\n",
            "Epoch 298/300\n",
            "1/1 - 0s - loss: 0.0012\n",
            "Epoch 299/300\n",
            "1/1 - 0s - loss: 0.0012\n",
            "Epoch 300/300\n",
            "1/1 - 0s - loss: 0.0012\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f61f639a390>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "2llS4XcdZnoo",
        "outputId": "ae300836-3a66-4caf-c017-b05835a5acf4"
      },
      "source": [
        "EPOCH_NUMBERS=[1,25,50,150,300] # Update me to see other Epochs\n",
        "plt.plot(xs,ys,label = \"Ys\")\n",
        "for EPOCH in EPOCH_NUMBERS:\n",
        "    plt.plot(xs,predictions[EPOCH-1],label = \"Epoch = \" + str(EPOCH))\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}