tiny_fltlnd.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "## Data import"
      ],
      "metadata": {
        "id": "ZL_hYOTs3XWB"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vMPKHD4rdsaK",
        "outputId": "6cb20ab9-c3fb-4b1e-e997-746da7e8d19e"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ]
        }
      ],
      "source": [
        "import numpy as np\n",
        "from google.colab import drive\n",
        "drive.mount('/content/drive')\n",
        "X_train = np.load('/content/drive/MyDrive/Colab Notebooks/Duomenų folderis/X_train.npy')\n",
        "X_pred = np.load('/content/drive/MyDrive/Colab Notebooks/Duomenų folderis/X_pred.npy')\n",
        "X_val = np.load('/content/drive/MyDrive/Colab Notebooks/Duomenų folderis/X_val.npy')\n",
        "y_val = np.load('/content/drive/MyDrive/Colab Notebooks/Duomenų folderis/y_val.npy')\n",
        "X_test = np.load('/content/drive/MyDrive/Colab Notebooks/Duomenų folderis/X_test.npy')\n",
        "y_test = np.load('/content/drive/MyDrive/Colab Notebooks/Duomenų folderis/y_test.npy')"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from fltlnd_imports import *"
      ],
      "metadata": {
        "id": "DoHljyLAejB2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Building a tiny model"
      ],
      "metadata": {
        "id": "8NXisoye3bmm"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "mimic_model = Sequential([\n",
        "    #MaxPooling2D((2,2)),\n",
        "    Conv2D(10, (3, 3), padding = 'same', activation='relu', kernel_initializer='he_uniform', strides=(2,2)),\n",
        "    Conv2D(10, (3, 3), padding = 'same', activation='relu', kernel_initializer='he_uniform', strides=(1,1)),\n",
        "    Conv2D(10, (3, 3), padding = 'same', activation='relu', kernel_initializer='he_uniform', strides=(1,1)),\n",
        "    Conv2D(10, (3, 3), padding = 'same', activation='relu', kernel_initializer='he_uniform', strides=(1,1)),\n",
        "    Conv2D(10, (3, 3), padding = 'same', activation='relu', kernel_initializer='he_uniform', strides=(2,2)),\n",
        "    Conv2D(10, (3, 3), padding = 'same', activation='relu', kernel_initializer='he_uniform', strides=(1,1)),\n",
        "    Conv2D(10, (3, 3), padding = 'same', activation='relu', kernel_initializer='he_uniform', strides=(1,1)),\n",
        "    Conv2D(10, (3, 3), padding = 'same', activation='relu', kernel_initializer='he_uniform', strides=(1,1)),\n",
        "    Conv2D(10, (3, 3), padding = 'same', activation='relu', kernel_initializer='he_uniform', strides=(2,2)),\n",
        "    Conv2D(10, (3, 3), padding = 'same', activation='relu', kernel_initializer='he_uniform', strides=(1,1)),\n",
        "    Conv2D(10, (3, 3), padding = 'same', activation='relu', kernel_initializer='he_uniform', strides=(1,1)),\n",
        "    Conv2D(10, (3, 3), padding = 'same', activation='relu', kernel_initializer='he_uniform', strides=(1,1)),\n",
        "    Flatten(),\n",
        "    Dense(30, activation='relu'),\n",
        "    Dropout(0.3),\n",
        "    Dense(5, activation='softmax')\n",
        "])\n",
        "mimic_model.build((None, 50,50,1))\n",
        "mimic_model.compile(loss='categorical_crossentropy',\n",
        "              optimizer=Adam(learning_rate=3e-4))"
      ],
      "metadata": {
        "id": "rVVIQQD-e3GH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "mimic_model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9RLGStwNhFs2",
        "outputId": "ed76e777-9e1d-4a2b-faeb-f5beb01d4ab9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_16\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d_68 (Conv2D)          (None, 25, 25, 10)        100       \n",
            "                                                                 \n",
            " conv2d_69 (Conv2D)          (None, 25, 25, 10)        910       \n",
            "                                                                 \n",
            " conv2d_70 (Conv2D)          (None, 25, 25, 10)        910       \n",
            "                                                                 \n",
            " conv2d_71 (Conv2D)          (None, 25, 25, 10)        910       \n",
            "                                                                 \n",
            " conv2d_72 (Conv2D)          (None, 13, 13, 10)        910       \n",
            "                                                                 \n",
            " conv2d_73 (Conv2D)          (None, 13, 13, 10)        910       \n",
            "                                                                 \n",
            " conv2d_74 (Conv2D)          (None, 13, 13, 10)        910       \n",
            "                                                                 \n",
            " conv2d_75 (Conv2D)          (None, 13, 13, 10)        910       \n",
            "                                                                 \n",
            " conv2d_76 (Conv2D)          (None, 7, 7, 10)          910       \n",
            "                                                                 \n",
            " conv2d_77 (Conv2D)          (None, 7, 7, 10)          910       \n",
            "                                                                 \n",
            " conv2d_78 (Conv2D)          (None, 7, 7, 10)          910       \n",
            "                                                                 \n",
            " conv2d_79 (Conv2D)          (None, 7, 7, 10)          910       \n",
            "                                                                 \n",
            " flatten_16 (Flatten)        (None, 490)               0         \n",
            "                                                                 \n",
            " dense_47 (Dense)            (None, 30)                14730     \n",
            "                                                                 \n",
            " dropout_30 (Dropout)        (None, 30)                0         \n",
            "                                                                 \n",
            " dense_48 (Dense)            (None, 5)                 155       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 24995 (97.64 KB)\n",
            "Trainable params: 24995 (97.64 KB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Original model:** \\\\\n",
        "loss: [0.36], val_loss: [0.11]\n",
        "\n",
        "Train set accuracy: 0.9667857 \\\\\n",
        "Validation set accuracy: 0.9845"
      ],
      "metadata": {
        "id": "ARy_LkJzoHGa"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "train_and_eval_pipe(mimic_model, X_train, X_pred, X_val, y_val, epochs_number = 350)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 507
        },
        "id": "tZOg7dcphGyM",
        "outputId": "33622b23-b169-47bb-9dbb-57ea85e18bc9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train set accuracy: 0.9665476\n",
            "Validation set accuracy: 0.984\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Original model:** \\\\\n",
        "Test set accuracy: 0.9825"
      ],
      "metadata": {
        "id": "-GyfOwtJq0zj"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "test_set_accuracy(mimic_model, X_test, y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4m8P71TlkfEl",
        "outputId": "cb929063-a8c9-4b7c-d974-8a3bfe988529"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Test set accuracy: 0.975\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "mimic_model.save('model115_6.h5')"
      ],
      "metadata": {
        "id": "gzAu5VuKqx-V"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}