Layers_Acc.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "HighLayers.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "DDMso7C_BZpX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "\n",
        "from __future__ import absolute_import\n",
        "from __future__ import division\n",
        "from __future__ import print_function\n",
        "\n",
        "from keras.engine.base_layer import Layer\n",
        "from keras.layers import Activation, Dense\n",
        "from keras import backend as K\n",
        "from sklearn.model_selection import train_test_split\n",
        "from keras.datasets import mnist\n",
        "from keras.optimizers import SGD\n",
        "from keras.utils import np_utils\n",
        "from __future__ import print_function\n",
        "import keras\n",
        "from keras.models import Sequential\n",
        "from keras.layers.core import Flatten\n",
        "from keras.layers import Dropout\n",
        "from keras.layers import Conv2D, MaxPooling2D\n",
        "from keras.layers.normalization import BatchNormalization\n",
        "import numpy as np\n",
        "\n",
        "class Mish(Layer):\n",
        "    '''\n",
        "    Mish Activation Function.\n",
        "    .. math::\n",
        "        mish(x) = x * tanh(softplus(x)) = x * tanh(ln(1 + e^{x}))\n",
        "    Shape:\n",
        "        - Input: Arbitrary. Use the keyword argument `input_shape`\n",
        "        (tuple of integers, does not include the samples axis)\n",
        "        when using this layer as the first layer in a model.\n",
        "        - Output: Same shape as the input.\n",
        "    Examples:\n",
        "        >>> X_input = Input(input_shape)\n",
        "        >>> X = Mish()(X_input)\n",
        "    '''\n",
        "\n",
        "    def __init__(self, **kwargs):\n",
        "        super(Mish, self).__init__(**kwargs)\n",
        "        self.supports_masking = True\n",
        "\n",
        "    def call(self, inputs):\n",
        "        return inputs * K.tanh(K.softplus(inputs))\n",
        "\n",
        "    def get_config(self):\n",
        "        base_config = super(Mish, self).get_config()\n",
        "        return dict(list(base_config.items()) + list(config.items()))\n",
        "\n",
        "    def compute_output_shape(self, input_shape):\n",
        "        return input_shape"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9f5JVaq3jjnM",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def mish(x):\n",
        "\treturn keras.layers.Lambda(lambda x: x*K.tanh(K.softplus(x)))(x)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "odZlsbUxCpnU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "4c1fd59d-c865-4527-f7d8-d718eeeca349"
      },
      "source": [
        "batch_size = 128\n",
        "num_classes = 10\n",
        "epochs = 10\n",
        "\n",
        "# input image dimensions\n",
        "img_rows, img_cols = 28, 28\n",
        "\n",
        "# the data, split between train and test sets\n",
        "(x_train, y_train), (x_test, y_test) = mnist.load_data()\n",
        "\n",
        "if K.image_data_format() == 'channels_first':\n",
        "    x_train = x_train.reshape(x_train.shape[0], 1, img_rows, img_cols)\n",
        "    x_test = x_test.reshape(x_test.shape[0], 1, img_rows, img_cols)\n",
        "    input_shape = (1, img_rows, img_cols)\n",
        "else:\n",
        "    x_train = x_train.reshape(x_train.shape[0], img_rows, img_cols, 1)\n",
        "    x_test = x_test.reshape(x_test.shape[0], img_rows, img_cols, 1)\n",
        "    input_shape = (img_rows, img_cols, 1)\n",
        "\n",
        "x_train = x_train.astype('float32')\n",
        "x_test = x_test.astype('float32')\n",
        "x_train /= 255\n",
        "x_test /= 255\n",
        "print('x_train shape:', x_train.shape)\n",
        "print(x_train.shape[0], 'train samples')\n",
        "print(x_test.shape[0], 'test samples')\n",
        "\n",
        "# convert class vectors to binary class matrices\n",
        "y_train = keras.utils.to_categorical(y_train, num_classes)\n",
        "y_test = keras.utils.to_categorical(y_test, num_classes)\n",
        "\n",
        "\n",
        "def model_relu(num_layers, x_train, y_train, batch_size, epochs, x_test, y_test):\n",
        "  model = Sequential()\n",
        "  model.add(Conv2D(20, kernel_size=(5, 5),\n",
        "                   activation=mish,\n",
        "                   input_shape=input_shape))\n",
        "  model.add(Conv2D(50, (5, 5), activation=mish))\n",
        "  model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "  model.add(Dropout(0.25))\n",
        "  model.add(Flatten())\n",
        "  for layers in range(num_layers):\n",
        "    model.add(Dense(500))\n",
        "    model.add(BatchNormalization())\n",
        "    model.add(Mish())\n",
        "    model.add(Dropout(0.25))\n",
        "  model.add(Dense(num_classes, activation='softmax'))\n",
        "\n",
        "  model.compile(loss=keras.losses.categorical_crossentropy,\n",
        "                optimizer=keras.optimizers.SGD(),\n",
        "                metrics=['accuracy'])\n",
        "\n",
        "  model.fit(x_train, y_train,\n",
        "            batch_size=batch_size,\n",
        "            epochs=epochs,\n",
        "            verbose=1,\n",
        "            validation_data=(x_test, y_test))\n",
        "  score = model.evaluate(x_test, y_test, verbose=0)\n",
        "  return score[1]\n",
        "\n",
        "l1 = []\n",
        "for x in range(12,23):\n",
        "  test_acc = model_relu(x,  x_train, y_train, batch_size, epochs, x_test, y_test)\n",
        "  l1.append(test_acc)"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "x_train shape: (60000, 28, 28, 1)\n",
            "60000 train samples\n",
            "10000 test samples\n",
            "Train on 60000 samples, validate on 10000 samples\n",
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 61s 1ms/step - loss: 1.5130 - acc: 0.4793 - val_loss: 0.4878 - val_acc: 0.8594\n",
            "Epoch 2/10\n",
            "60000/60000 [==============================] - 28s 471us/step - loss: 0.5191 - acc: 0.8393 - val_loss: 0.2602 - val_acc: 0.9230\n",
            "Epoch 3/10\n",
            "60000/60000 [==============================] - 29s 477us/step - loss: 0.3460 - acc: 0.8968 - val_loss: 0.1868 - val_acc: 0.9442\n",
            "Epoch 4/10\n",
            "60000/60000 [==============================] - 28s 468us/step - loss: 0.2674 - acc: 0.9204 - val_loss: 0.1227 - val_acc: 0.9631\n",
            "Epoch 5/10\n",
            "60000/60000 [==============================] - 28s 468us/step - loss: 0.2164 - acc: 0.9378 - val_loss: 0.1030 - val_acc: 0.9672\n",
            "Epoch 6/10\n",
            "60000/60000 [==============================] - 28s 466us/step - loss: 0.1858 - acc: 0.9464 - val_loss: 0.0852 - val_acc: 0.9746\n",
            "Epoch 7/10\n",
            "60000/60000 [==============================] - 28s 465us/step - loss: 0.1645 - acc: 0.9520 - val_loss: 0.0746 - val_acc: 0.9771\n",
            "Epoch 8/10\n",
            "60000/60000 [==============================] - 28s 468us/step - loss: 0.1496 - acc: 0.9564 - val_loss: 0.0670 - val_acc: 0.9798\n",
            "Epoch 9/10\n",
            "60000/60000 [==============================] - 28s 467us/step - loss: 0.1334 - acc: 0.9606 - val_loss: 0.0624 - val_acc: 0.9813\n",
            "Epoch 10/10\n",
            "60000/60000 [==============================] - 28s 473us/step - loss: 0.1237 - acc: 0.9645 - val_loss: 0.0562 - val_acc: 0.9843\n",
            "Train on 60000 samples, validate on 10000 samples\n",
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 64s 1ms/step - loss: 1.7412 - acc: 0.3943 - val_loss: 0.8820 - val_acc: 0.7148\n",
            "Epoch 2/10\n",
            "60000/60000 [==============================] - 30s 497us/step - loss: 0.6332 - acc: 0.8008 - val_loss: 0.3446 - val_acc: 0.8956\n",
            "Epoch 3/10\n",
            "60000/60000 [==============================] - 30s 497us/step - loss: 0.3862 - acc: 0.8836 - val_loss: 0.2058 - val_acc: 0.9373\n",
            "Epoch 4/10\n",
            "60000/60000 [==============================] - 30s 492us/step - loss: 0.2863 - acc: 0.9152 - val_loss: 0.1370 - val_acc: 0.9586\n",
            "Epoch 5/10\n",
            "60000/60000 [==============================] - 30s 492us/step - loss: 0.2274 - acc: 0.9342 - val_loss: 0.1063 - val_acc: 0.9683\n",
            "Epoch 6/10\n",
            "60000/60000 [==============================] - 29s 491us/step - loss: 0.1891 - acc: 0.9446 - val_loss: 0.0860 - val_acc: 0.9738\n",
            "Epoch 7/10\n",
            "60000/60000 [==============================] - 29s 490us/step - loss: 0.1679 - acc: 0.9510 - val_loss: 0.0793 - val_acc: 0.9746\n",
            "Epoch 8/10\n",
            "60000/60000 [==============================] - 29s 490us/step - loss: 0.1521 - acc: 0.9573 - val_loss: 0.0625 - val_acc: 0.9813\n",
            "Epoch 9/10\n",
            "60000/60000 [==============================] - 30s 496us/step - loss: 0.1363 - acc: 0.9603 - val_loss: 0.0591 - val_acc: 0.9814\n",
            "Epoch 10/10\n",
            "60000/60000 [==============================] - 30s 496us/step - loss: 0.1236 - acc: 0.9644 - val_loss: 0.0538 - val_acc: 0.9828\n",
            "Train on 60000 samples, validate on 10000 samples\n",
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 65s 1ms/step - loss: 1.8932 - acc: 0.3192 - val_loss: 1.1149 - val_acc: 0.6492\n",
            "Epoch 2/10\n",
            "60000/60000 [==============================] - 31s 514us/step - loss: 0.7889 - acc: 0.7403 - val_loss: 0.4223 - val_acc: 0.8767\n",
            "Epoch 3/10\n",
            "60000/60000 [==============================] - 31s 513us/step - loss: 0.4400 - acc: 0.8675 - val_loss: 0.2463 - val_acc: 0.9269\n",
            "Epoch 4/10\n",
            "60000/60000 [==============================] - 31s 514us/step - loss: 0.3107 - acc: 0.9073 - val_loss: 0.1686 - val_acc: 0.9483\n",
            "Epoch 5/10\n",
            "60000/60000 [==============================] - 31s 516us/step - loss: 0.2473 - acc: 0.9287 - val_loss: 0.1083 - val_acc: 0.9670\n",
            "Epoch 6/10\n",
            "60000/60000 [==============================] - 31s 512us/step - loss: 0.2011 - acc: 0.9420 - val_loss: 0.0968 - val_acc: 0.9723\n",
            "Epoch 7/10\n",
            "60000/60000 [==============================] - 31s 517us/step - loss: 0.1738 - acc: 0.9512 - val_loss: 0.0799 - val_acc: 0.9751\n",
            "Epoch 8/10\n",
            "60000/60000 [==============================] - 31s 525us/step - loss: 0.1563 - acc: 0.9557 - val_loss: 0.0687 - val_acc: 0.9791\n",
            "Epoch 9/10\n",
            "60000/60000 [==============================] - 31s 512us/step - loss: 0.1406 - acc: 0.9598 - val_loss: 0.0577 - val_acc: 0.9826\n",
            "Epoch 10/10\n",
            "60000/60000 [==============================] - 31s 514us/step - loss: 0.1346 - acc: 0.9618 - val_loss: 0.0552 - val_acc: 0.9838\n",
            "Train on 60000 samples, validate on 10000 samples\n",
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 68s 1ms/step - loss: 2.0126 - acc: 0.2809 - val_loss: 2.0061 - val_acc: 0.2551\n",
            "Epoch 2/10\n",
            "60000/60000 [==============================] - 32s 539us/step - loss: 1.0212 - acc: 0.6298 - val_loss: 0.7946 - val_acc: 0.7402\n",
            "Epoch 3/10\n",
            "60000/60000 [==============================] - 32s 541us/step - loss: 0.5524 - acc: 0.8255 - val_loss: 0.3511 - val_acc: 0.8988\n",
            "Epoch 4/10\n",
            "60000/60000 [==============================] - 32s 537us/step - loss: 0.3723 - acc: 0.8874 - val_loss: 0.2396 - val_acc: 0.9288\n",
            "Epoch 5/10\n",
            "60000/60000 [==============================] - 32s 537us/step - loss: 0.2881 - acc: 0.9154 - val_loss: 0.1364 - val_acc: 0.9578\n",
            "Epoch 6/10\n",
            "60000/60000 [==============================] - 33s 546us/step - loss: 0.2366 - acc: 0.9315 - val_loss: 0.1239 - val_acc: 0.9613\n",
            "Epoch 7/10\n",
            "60000/60000 [==============================] - 32s 541us/step - loss: 0.2081 - acc: 0.9406 - val_loss: 0.0866 - val_acc: 0.9723\n",
            "Epoch 8/10\n",
            "60000/60000 [==============================] - 32s 538us/step - loss: 0.1814 - acc: 0.9496 - val_loss: 0.0765 - val_acc: 0.9776\n",
            "Epoch 9/10\n",
            "60000/60000 [==============================] - 32s 538us/step - loss: 0.1604 - acc: 0.9539 - val_loss: 0.0672 - val_acc: 0.9793\n",
            "Epoch 10/10\n",
            "60000/60000 [==============================] - 32s 540us/step - loss: 0.1469 - acc: 0.9586 - val_loss: 0.0628 - val_acc: 0.9824\n",
            "Train on 60000 samples, validate on 10000 samples\n",
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 70s 1ms/step - loss: 2.1108 - acc: 0.2355 - val_loss: 2.3297 - val_acc: 0.2194\n",
            "Epoch 2/10\n",
            "60000/60000 [==============================] - 34s 559us/step - loss: 1.2068 - acc: 0.5547 - val_loss: 0.8248 - val_acc: 0.7598\n",
            "Epoch 3/10\n",
            "60000/60000 [==============================] - 34s 560us/step - loss: 0.6690 - acc: 0.7796 - val_loss: 0.3669 - val_acc: 0.8989\n",
            "Epoch 4/10\n",
            "60000/60000 [==============================] - 34s 566us/step - loss: 0.4454 - acc: 0.8658 - val_loss: 0.2180 - val_acc: 0.9381\n",
            "Epoch 5/10\n",
            "60000/60000 [==============================] - 34s 570us/step - loss: 0.3284 - acc: 0.9048 - val_loss: 0.1577 - val_acc: 0.9543\n",
            "Epoch 6/10\n",
            "60000/60000 [==============================] - 34s 564us/step - loss: 0.2664 - acc: 0.9234 - val_loss: 0.1271 - val_acc: 0.9614\n",
            "Epoch 7/10\n",
            "60000/60000 [==============================] - 34s 559us/step - loss: 0.2248 - acc: 0.9382 - val_loss: 0.0997 - val_acc: 0.9707\n",
            "Epoch 8/10\n",
            "60000/60000 [==============================] - 33s 555us/step - loss: 0.1957 - acc: 0.9447 - val_loss: 0.0966 - val_acc: 0.9703\n",
            "Epoch 9/10\n",
            "60000/60000 [==============================] - 33s 554us/step - loss: 0.1796 - acc: 0.9499 - val_loss: 0.0792 - val_acc: 0.9755\n",
            "Epoch 10/10\n",
            "60000/60000 [==============================] - 34s 560us/step - loss: 0.1594 - acc: 0.9565 - val_loss: 0.0672 - val_acc: 0.9805\n",
            "Train on 60000 samples, validate on 10000 samples\n",
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 73s 1ms/step - loss: 2.1499 - acc: 0.2185 - val_loss: 2.6395 - val_acc: 0.1551\n",
            "Epoch 2/10\n",
            "60000/60000 [==============================] - 35s 580us/step - loss: 1.3889 - acc: 0.4773 - val_loss: 1.1772 - val_acc: 0.6296\n",
            "Epoch 3/10\n",
            "60000/60000 [==============================] - 35s 575us/step - loss: 0.7179 - acc: 0.7638 - val_loss: 0.4733 - val_acc: 0.8718\n",
            "Epoch 4/10\n",
            "60000/60000 [==============================] - 35s 576us/step - loss: 0.4674 - acc: 0.8596 - val_loss: 0.2741 - val_acc: 0.9248\n",
            "Epoch 5/10\n",
            "60000/60000 [==============================] - 35s 578us/step - loss: 0.3661 - acc: 0.8924 - val_loss: 0.1908 - val_acc: 0.9463\n",
            "Epoch 6/10\n",
            "60000/60000 [==============================] - 34s 574us/step - loss: 0.2941 - acc: 0.9149 - val_loss: 0.1498 - val_acc: 0.9574\n",
            "Epoch 7/10\n",
            "60000/60000 [==============================] - 35s 576us/step - loss: 0.2518 - acc: 0.9294 - val_loss: 0.1229 - val_acc: 0.9654\n",
            "Epoch 8/10\n",
            "60000/60000 [==============================] - 35s 587us/step - loss: 0.2233 - acc: 0.9375 - val_loss: 0.1047 - val_acc: 0.9701\n",
            "Epoch 9/10\n",
            "60000/60000 [==============================] - 35s 585us/step - loss: 0.2002 - acc: 0.9438 - val_loss: 0.0942 - val_acc: 0.9720\n",
            "Epoch 10/10\n",
            "60000/60000 [==============================] - 35s 578us/step - loss: 0.1817 - acc: 0.9490 - val_loss: 0.0832 - val_acc: 0.9752\n",
            "Train on 60000 samples, validate on 10000 samples\n",
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 75s 1ms/step - loss: 2.1595 - acc: 0.2101 - val_loss: 2.5745 - val_acc: 0.1141\n",
            "Epoch 2/10\n",
            "60000/60000 [==============================] - 36s 601us/step - loss: 1.7262 - acc: 0.3358 - val_loss: 1.9629 - val_acc: 0.2429\n",
            "Epoch 3/10\n",
            "60000/60000 [==============================] - 36s 601us/step - loss: 1.0502 - acc: 0.6171 - val_loss: 0.7043 - val_acc: 0.7784\n",
            "Epoch 4/10\n",
            "60000/60000 [==============================] - 36s 600us/step - loss: 0.6547 - acc: 0.7868 - val_loss: 0.3730 - val_acc: 0.8969\n",
            "Epoch 5/10\n",
            "60000/60000 [==============================] - 36s 601us/step - loss: 0.4680 - acc: 0.8585 - val_loss: 0.2289 - val_acc: 0.9385\n",
            "Epoch 6/10\n",
            "60000/60000 [==============================] - 36s 601us/step - loss: 0.3643 - acc: 0.8941 - val_loss: 0.1724 - val_acc: 0.9540\n",
            "Epoch 7/10\n",
            "60000/60000 [==============================] - 36s 606us/step - loss: 0.2934 - acc: 0.9151 - val_loss: 0.1300 - val_acc: 0.9656\n",
            "Epoch 8/10\n",
            "60000/60000 [==============================] - 36s 607us/step - loss: 0.2498 - acc: 0.9296 - val_loss: 0.1083 - val_acc: 0.9716\n",
            "Epoch 9/10\n",
            "60000/60000 [==============================] - 36s 599us/step - loss: 0.2193 - acc: 0.9399 - val_loss: 0.0986 - val_acc: 0.9730\n",
            "Epoch 10/10\n",
            "60000/60000 [==============================] - 36s 599us/step - loss: 0.1992 - acc: 0.9450 - val_loss: 0.0796 - val_acc: 0.9786\n",
            "Train on 60000 samples, validate on 10000 samples\n",
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 78s 1ms/step - loss: 2.1704 - acc: 0.2019 - val_loss: 2.7488 - val_acc: 0.1151\n",
            "Epoch 2/10\n",
            "60000/60000 [==============================] - 38s 631us/step - loss: 1.7385 - acc: 0.2962 - val_loss: 2.7440 - val_acc: 0.1465\n",
            "Epoch 3/10\n",
            "60000/60000 [==============================] - 38s 634us/step - loss: 1.4060 - acc: 0.4040 - val_loss: 1.4998 - val_acc: 0.3863\n",
            "Epoch 4/10\n",
            "60000/60000 [==============================] - 38s 626us/step - loss: 1.0669 - acc: 0.5590 - val_loss: 1.0369 - val_acc: 0.5638\n",
            "Epoch 5/10\n",
            "60000/60000 [==============================] - 38s 630us/step - loss: 0.7941 - acc: 0.7169 - val_loss: 0.5559 - val_acc: 0.8667\n",
            "Epoch 6/10\n",
            "60000/60000 [==============================] - 37s 622us/step - loss: 0.5979 - acc: 0.8095 - val_loss: 0.4299 - val_acc: 0.8890\n",
            "Epoch 7/10\n",
            "60000/60000 [==============================] - 37s 624us/step - loss: 0.4848 - acc: 0.8541 - val_loss: 0.3067 - val_acc: 0.9175\n",
            "Epoch 8/10\n",
            "60000/60000 [==============================] - 37s 622us/step - loss: 0.4102 - acc: 0.8814 - val_loss: 0.2327 - val_acc: 0.9398\n",
            "Epoch 9/10\n",
            "60000/60000 [==============================] - 37s 622us/step - loss: 0.3439 - acc: 0.9032 - val_loss: 0.1729 - val_acc: 0.9537\n",
            "Epoch 10/10\n",
            "60000/60000 [==============================] - 37s 621us/step - loss: 0.2986 - acc: 0.9176 - val_loss: 0.1322 - val_acc: 0.9646\n",
            "Train on 60000 samples, validate on 10000 samples\n",
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 80s 1ms/step - loss: 2.1675 - acc: 0.1986 - val_loss: 2.8450 - val_acc: 0.1135\n",
            "Epoch 2/10\n",
            "60000/60000 [==============================] - 39s 650us/step - loss: 1.7407 - acc: 0.2948 - val_loss: 2.9005 - val_acc: 0.1166\n",
            "Epoch 3/10\n",
            "60000/60000 [==============================] - 39s 645us/step - loss: 1.4595 - acc: 0.3861 - val_loss: 1.7066 - val_acc: 0.3014\n",
            "Epoch 4/10\n",
            "60000/60000 [==============================] - 39s 643us/step - loss: 1.1691 - acc: 0.5291 - val_loss: 1.0062 - val_acc: 0.6519\n",
            "Epoch 5/10\n",
            "60000/60000 [==============================] - 39s 643us/step - loss: 0.8199 - acc: 0.7152 - val_loss: 0.7497 - val_acc: 0.7696\n",
            "Epoch 6/10\n",
            "60000/60000 [==============================] - 39s 649us/step - loss: 0.6368 - acc: 0.7939 - val_loss: 0.5987 - val_acc: 0.8246\n",
            "Epoch 7/10\n",
            "60000/60000 [==============================] - 39s 652us/step - loss: 0.5268 - acc: 0.8360 - val_loss: 0.4609 - val_acc: 0.8664\n",
            "Epoch 8/10\n",
            "60000/60000 [==============================] - 40s 659us/step - loss: 0.4509 - acc: 0.8644 - val_loss: 0.3419 - val_acc: 0.9052\n",
            "Epoch 9/10\n",
            "60000/60000 [==============================] - 39s 644us/step - loss: 0.3838 - acc: 0.8883 - val_loss: 0.2486 - val_acc: 0.9343\n",
            "Epoch 10/10\n",
            "60000/60000 [==============================] - 39s 649us/step - loss: 0.3405 - acc: 0.9024 - val_loss: 0.2006 - val_acc: 0.9452\n",
            "Train on 60000 samples, validate on 10000 samples\n",
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 88s 1ms/step - loss: 2.1900 - acc: 0.1983 - val_loss: 2.6362 - val_acc: 0.1135\n",
            "Epoch 2/10\n",
            "60000/60000 [==============================] - 40s 672us/step - loss: 1.8457 - acc: 0.2693 - val_loss: 2.6581 - val_acc: 0.1068\n",
            "Epoch 3/10\n",
            "60000/60000 [==============================] - 41s 682us/step - loss: 1.6116 - acc: 0.3306 - val_loss: 1.9764 - val_acc: 0.1862\n",
            "Epoch 4/10\n",
            "60000/60000 [==============================] - 40s 673us/step - loss: 1.3877 - acc: 0.3938 - val_loss: 1.5774 - val_acc: 0.3504\n",
            "Epoch 5/10\n",
            "60000/60000 [==============================] - 40s 669us/step - loss: 1.2384 - acc: 0.4705 - val_loss: 1.2824 - val_acc: 0.5124\n",
            "Epoch 6/10\n",
            "60000/60000 [==============================] - 40s 673us/step - loss: 1.0373 - acc: 0.5876 - val_loss: 0.9611 - val_acc: 0.6578\n",
            "Epoch 7/10\n",
            "60000/60000 [==============================] - 40s 675us/step - loss: 0.8252 - acc: 0.7107 - val_loss: 0.5903 - val_acc: 0.8195\n",
            "Epoch 8/10\n",
            "60000/60000 [==============================] - 40s 672us/step - loss: 0.5975 - acc: 0.8149 - val_loss: 0.3749 - val_acc: 0.8943\n",
            "Epoch 9/10\n",
            "60000/60000 [==============================] - 40s 674us/step - loss: 0.4730 - acc: 0.8601 - val_loss: 0.2711 - val_acc: 0.9314\n",
            "Epoch 10/10\n",
            "60000/60000 [==============================] - 40s 674us/step - loss: 0.3975 - acc: 0.8867 - val_loss: 0.2158 - val_acc: 0.9466\n",
            "Train on 60000 samples, validate on 10000 samples\n",
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 87s 1ms/step - loss: 2.1826 - acc: 0.2011 - val_loss: 2.6562 - val_acc: 0.1135\n",
            "Epoch 2/10\n",
            "60000/60000 [==============================] - 42s 705us/step - loss: 1.8207 - acc: 0.2802 - val_loss: 3.0680 - val_acc: 0.1138\n",
            "Epoch 3/10\n",
            "60000/60000 [==============================] - 43s 714us/step - loss: 1.5430 - acc: 0.3480 - val_loss: 2.2197 - val_acc: 0.1590\n",
            "Epoch 4/10\n",
            "60000/60000 [==============================] - 43s 712us/step - loss: 1.3665 - acc: 0.4211 - val_loss: 1.7220 - val_acc: 0.3014\n",
            "Epoch 5/10\n",
            "60000/60000 [==============================] - 43s 714us/step - loss: 1.1674 - acc: 0.5241 - val_loss: 1.5301 - val_acc: 0.4196\n",
            "Epoch 6/10\n",
            "60000/60000 [==============================] - 43s 709us/step - loss: 1.0365 - acc: 0.5847 - val_loss: 1.1859 - val_acc: 0.5456\n",
            "Epoch 7/10\n",
            "60000/60000 [==============================] - 43s 713us/step - loss: 0.9209 - acc: 0.6336 - val_loss: 0.9219 - val_acc: 0.6780\n",
            "Epoch 8/10\n",
            "60000/60000 [==============================] - 43s 713us/step - loss: 0.7890 - acc: 0.6954 - val_loss: 0.6518 - val_acc: 0.7962\n",
            "Epoch 9/10\n",
            "60000/60000 [==============================] - 43s 719us/step - loss: 0.6695 - acc: 0.7552 - val_loss: 0.5927 - val_acc: 0.7736\n",
            "Epoch 10/10\n",
            "60000/60000 [==============================] - 43s 709us/step - loss: 0.5725 - acc: 0.8016 - val_loss: 0.4906 - val_acc: 0.8516\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Zl0JJxxt5kx_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "l2 = [0.982,\n",
        " 0.9791,\n",
        " 0.9795,\n",
        " 0.9752,\n",
        " 0.972,\n",
        " 0.9622,\n",
        " 0.9557,\n",
        " 0.9174,\n",
        " 0.7358,\n",
        " 0.6341,\n",
        " 0.3968]\n",
        "l2 = [i * 100 for i in l2]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dKKizHzm1Jek",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 749
        },
        "outputId": "632e9bdb-2cee-4a6c-80f7-f31bdcfacb54"
      },
      "source": [
        "positions = (0, 2, 4, 6, 8, 10)\n",
        "labels = (15, 17, 19, 21, 23, 25)\n",
        "\n",
        "fig = plt.figure(figsize=(12,12))\n",
        "ax = fig.add_subplot(111)\n",
        "plt.plot(l,'b-o', label= \"ReLU\")\n",
        "plt.plot(l1, 'r-o', label= \"Mish\")\n",
        "plt.plot(l2, 'k-o', label= \"Swish\")\n",
        "plt.legend(loc='best', fontsize=15)\n",
        "plt.xticks(positions, labels)\n",
        "plt.grid()\n",
        "plt.xlabel('Number of Layers', fontsize = 20)\n",
        "plt.ylabel('Testing Accuracy', fontsize = 20)\n",
        "ax.tick_params(axis='both', which='major', labelsize=15)\n",
        "ax.title.set_text('Testing Accuracy vs Number of Layers on MNIST')\n",
        "ax.title.set_fontsize(20)\n",
        "plt.savefig(\"layersacc.png\", bbox_inches = 'tight')\n",
        "plt.show()"
      ],
      "execution_count": 63,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAuoAAALcCAYAAACxRJUWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 864x864 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r579LKLr2feF",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}