Upload Зачетное задание.ipynb

Зачетное задание.ipynb

@@ -0,0 +1,292 @@
+{
+  "nbformat": 4,
+  "nbformat_minor": 0,
+  "metadata": {
+    "colab": {
+      "provenance": []
+    },
+    "kernelspec": {
+      "name": "python3",
+      "display_name": "Python 3"
+    },
+    "language_info": {
+      "name": "python"
+    }
+  },
+  "cells": [
+    {
+      "cell_type": "code",
+      "execution_count": 80,
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "i5iQu5HRIhPG",
+        "outputId": "bfd92ab2-ecbe-4f7d-db79-ba0c6f859ec5"
+      },
+      "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 tensorflow as tf\n",
+        "from tensorflow import keras\n",
+        "from tensorflow.keras import layers\n",
+        "import numpy as np\n",
+        "import matplotlib.pyplot as plt\n",
+        "from tensorflow.keras.utils import plot_model\n",
+        "from sklearn.model_selection import train_test_split\n",
+        "from google.colab import drive\n",
+        "drive.mount('/content/drive')"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "(x_train, y_train), (x_test, y_test) = keras.datasets.mnist.load_data()\n",
+        "x_train, x_val, y_train, y_val = train_test_split(x_train, y_train, test_size=0.2, random_state=42)\n",
+        "len(x_train), len(y_train), len(x_test), len(y_test), len(x_val), len(y_val)"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "meXoyyAyQc9_",
+        "outputId": "60731583-6e4d-4dc8-b9c1-4aacfa688aa0"
+      },
+      "execution_count": 81,
+      "outputs": [
+        {
+          "output_type": "execute_result",
+          "data": {
+            "text/plain": [
+              "(48000, 48000, 10000, 10000, 12000, 12000)"
+            ]
+          },
+          "metadata": {},
+          "execution_count": 81
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "x_train = x_train / 255\n",
+        "x_test = x_test / 255\n",
+        "\n",
+        "y_train = y_train % 2\n",
+        "y_test = y_test % 2"
+      ],
+      "metadata": {
+        "id": "z5-ANJh9QdAk"
+      },
+      "execution_count": 82,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "model = keras.Sequential([\n",
+        "    layers.Flatten(input_shape=(28, 28)),\n",
+        "    layers.Dense(128, activation=\"relu\"),\n",
+        "    layers.Dense(1, activation=\"sigmoid\")\n",
+        "])\n",
+        "plot_model(model, to_file=\"model_architecture.png\", show_shapes=True)"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 422
+        },
+        "id": "_zDHmhqOQdC3",
+        "outputId": "17755be5-fef2-4023-bcb6-0468d10b98f0"
+      },
+      "execution_count": 92,
+      "outputs": [
+        {
+          "output_type": "execute_result",
+          "data": {
+            "image/png": "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\n",
+            "text/plain": [
+              "<IPython.core.display.Image object>"
+            ]
+          },
+          "metadata": {},
+          "execution_count": 92
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "model.compile(optimizer=\"adam\", loss=\"binary_crossentropy\", metrics=[\"accuracy\"])\n",
+        "\n",
+        "history = model.fit(x_train, y_train, batch_size=64, epochs=10, validation_data=(x_val, y_val))\n",
+        "\n",
+        "model.save(\"mnist_model.h5\")"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "Huhram2NQdGY",
+        "outputId": "09833bf3-d532-4857-f0b3-bf9e6977fbbf"
+      },
+      "execution_count": 93,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Epoch 1/10\n",
+            "750/750 [==============================] - 5s 5ms/step - loss: 0.1480 - accuracy: 0.9464 - val_loss: -337.0152 - val_accuracy: 0.2062\n",
+            "Epoch 2/10\n",
+            "750/750 [==============================] - 3s 4ms/step - loss: 0.0696 - accuracy: 0.9763 - val_loss: -144.5670 - val_accuracy: 0.2066\n",
+            "Epoch 3/10\n",
+            "750/750 [==============================] - 3s 4ms/step - loss: 0.0507 - accuracy: 0.9824 - val_loss: -471.2610 - val_accuracy: 0.2071\n",
+            "Epoch 4/10\n",
+            "750/750 [==============================] - 4s 5ms/step - loss: 0.0407 - accuracy: 0.9859 - val_loss: -629.9583 - val_accuracy: 0.2068\n",
+            "Epoch 5/10\n",
+            "750/750 [==============================] - 4s 6ms/step - loss: 0.0315 - accuracy: 0.9896 - val_loss: -1320.8126 - val_accuracy: 0.2069\n",
+            "Epoch 6/10\n",
+            "750/750 [==============================] - 3s 4ms/step - loss: 0.0254 - accuracy: 0.9914 - val_loss: -1551.2062 - val_accuracy: 0.2071\n",
+            "Epoch 7/10\n",
+            "750/750 [==============================] - 3s 4ms/step - loss: 0.0206 - accuracy: 0.9933 - val_loss: -2237.5056 - val_accuracy: 0.2070\n",
+            "Epoch 8/10\n",
+            "750/750 [==============================] - 4s 5ms/step - loss: 0.0171 - accuracy: 0.9948 - val_loss: -2575.4690 - val_accuracy: 0.2068\n",
+            "Epoch 9/10\n",
+            "750/750 [==============================] - 4s 5ms/step - loss: 0.0129 - accuracy: 0.9964 - val_loss: -1416.2383 - val_accuracy: 0.2069\n",
+            "Epoch 10/10\n",
+            "750/750 [==============================] - 3s 4ms/step - loss: 0.0121 - accuracy: 0.9967 - val_loss: -2124.9729 - val_accuracy: 0.2072\n"
+          ]
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "train_loss, train_acc = model.evaluate(x_train, y_train, verbose=0)\n",
+        "val_loss, val_acc = model.evaluate(x_val, y_val, verbose=0)\n",
+        "test_loss, test_acc = model.evaluate(x_test, y_test, verbose=0)\n",
+        "\n",
+        "print(\"Результаты обучения:\")\n",
+        "print(\"Тренировочная выборка - Loss:\", train_loss, \"Accuracy:\", train_acc)\n",
+        "print(\"Валидационная выборка - Loss:\", val_loss, \"Accuracy:\", val_acc)\n",
+        "print(\"Тестовая выборка - Loss:\", test_loss, \"Accuracy:\", test_acc)"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "nii8pfWCb-t3",
+        "outputId": "30624b5d-fed2-4d49-d4d2-eac5c801c6b7"
+      },
+      "execution_count": 94,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Результаты обучения:\n",
+            "Тренировочная выборка - Loss: 0.007834645919501781 Accuracy: 0.9984999895095825\n",
+            "Валидационная выборка - Loss: -2124.973388671875 Accuracy: 0.20724999904632568\n",
+            "Тестовая выборка - Loss: 0.039042405784130096 Accuracy: 0.988099992275238\n"
+          ]
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "predictions = model.predict(x_test)\n",
+        "predicted_labels = np.round(predictions).flatten()"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "oZPnJ2kpQ5YL",
+        "outputId": "7ec4fa91-0af3-4c62-fbba-4894ca1b32f7"
+      },
+      "execution_count": 95,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "313/313 [==============================] - 0s 1ms/step\n"
+          ]
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "random_indices = np.random.choice(len(x_test), 10, replace=False)\n",
+        "\n",
+        "plt.figure(figsize=(12, 8))\n",
+        "for i, idx in enumerate(random_indices):\n",
+        "    original_label = y_test[idx]\n",
+        "    predicted_label = predicted_labels[idx]\n",
+        "    original_image = x_test[idx]\n",
+        "\n",
+        "    plt.subplot(2, 5, i + 1)\n",
+        "    plt.imshow(original_image, cmap=\"gray\")\n",
+        "    plt.axis(\"off\")\n",
+        "    plt.title(f\"Фактическое: {original_label}, Предсказанное: {int(predicted_label)}\", fontsize=9)\n",
+        "\n",
+        "plt.tight_layout()\n",
+        "plt.show()"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 619
+        },
+        "id": "Z8wJZ-YlQ5bk",
+        "outputId": "0d3b1b3d-35a5-4d8e-f0c9-d47793869c31"
+      },
+      "execution_count": 96,
+      "outputs": [
+        {
+          "output_type": "display_data",
+          "data": {
+            "text/plain": [
+              "<Figure size 1200x800 with 10 Axes>"
+            ],
+            "image/png": "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\n"
+          },
+          "metadata": {}
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "model.save('drive/MyDrive/my_model')"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "x3XBk-u3RM_h",
+        "outputId": "9112d822-04dd-4365-cd30-d747e6609f45"
+      },
+      "execution_count": 97,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stderr",
+          "text": [
+            "WARNING:absl:Found untraced functions such as _update_step_xla while saving (showing 1 of 1). These functions will not be directly callable after loading.\n"
+          ]
+        }
+      ]
+    }
+  ]
+}