neuralnetwork.ipynb

{
  "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": 1,
      "metadata": {
        "id": "sakSwPyERrdG"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "from matplotlib import pyplot as plt\n",
        "import random\n",
        "from tqdm import tqdm\n",
        "from model.neuralMNIST import NeuralNetwork\n",
        "from model.neuralMNIST import CategoryEntropy\n",
        "from model.neuralMNIST import sparse_to_one_hot"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_data = pd.read_csv(\"/content/sample_data/mnist_train_small.csv\")\n",
        "train_images = train_data.iloc[:, 1:].values    #   shape(19999, 1, 28, 28)\n",
        "train_labels = train_data.iloc[:, 0].values\n",
        "print(train_images.shape)\n",
        "print(train_labels[0])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "M51D3HztTzdP",
        "outputId": "aaaf089a-a3bd-4bf4-9304-04840b48594b"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(19999, 784)\n",
            "5\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_datanlabel = pd.read_csv(\"/content/sample_data/mnist_test.csv\")\n",
        "test_images = test_datanlabel.iloc[:, 1:].values    #   shape(19999, 1, 28, 28)\n",
        "test_labels = test_datanlabel.iloc[:, 0].values\n",
        "\n",
        "print(test_images.shape)\n",
        "print(test_labels[0])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aioxyQnIW4tw",
        "outputId": "812a71fc-dcc7-486e-8a06-bab062cbebb5"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(9999, 784)\n",
            "2\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "myNeuralNet = NeuralNetwork(input_size=28*28, hidden_layers=[512, 512], output_size=10)"
      ],
      "metadata": {
        "id": "7Ep5JARoT5hc"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "data = train_images\n",
        "datalabels = train_labels\n",
        "\n",
        "data = (data.astype(np.float32)-127.5) / 127.5\n",
        "accuracies = []\n",
        "losses = []\n",
        "\n",
        "BATCH_SIZE = 32\n",
        "\n",
        "\n",
        "for epoch in range(1, 10):\n",
        "    train_steps = len(data) // BATCH_SIZE\n",
        "\n",
        "    for step in tqdm(range(train_steps), desc=f'Epoch {epoch}'):\n",
        "        batch_X = data[step * BATCH_SIZE:(step + 1) * BATCH_SIZE]\n",
        "        batch_y = datalabels[step * BATCH_SIZE:(step + 1) * BATCH_SIZE]\n",
        "\n",
        "        X = batch_X\n",
        "        y = batch_y\n",
        "        output = myNeuralNet.forward(X)\n",
        "\n",
        "        if step % 100 == 0:\n",
        "            predictions = np.argmax(output, axis=1)\n",
        "            if len(y.shape) == 2:\n",
        "                y = np.argmax(y, axis=1)\n",
        "            accuracy = np.mean(predictions == y)\n",
        "            loss = CategoryEntropy(output, sparse_to_one_hot(y, 10))\n",
        "            accuracies.append(accuracy)\n",
        "            losses.append(loss)\n",
        "\n",
        "            print(f'epoch:{epoch}, acc: {accuracy:.3f}, loss: {loss:.3f}')\n",
        "\n",
        "        myNeuralNet.backwards(y)\n",
        "        myNeuralNet.updateParams(lr=0.01, decay=1e-2)\n",
        "        myNeuralNet.gradientsWeights = []\n",
        "        myNeuralNet.gradientsBiases = []"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kIfp255tUAEf",
        "outputId": "6797a0c1-1eab-4357-8b66-a574898fac9a"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 1:   1%|          | 7/624 [00:00<00:09, 64.79it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:1, acc: 0.188, loss: 2.294\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 1:  18%|█▊        | 112/624 [00:01<00:06, 81.40it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:1, acc: 0.312, loss: 2.208\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 1:  34%|███▍      | 214/624 [00:02<00:04, 85.21it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:1, acc: 0.469, loss: 2.139\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 1:  51%|█████     | 316/624 [00:03<00:03, 95.49it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:1, acc: 0.531, loss: 2.039\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 1:  67%|██████▋   | 420/624 [00:04<00:02, 98.03it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:1, acc: 0.625, loss: 2.017\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 1:  82%|████████▏ | 513/624 [00:05<00:01, 97.41it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:1, acc: 0.688, loss: 1.864\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 1:  97%|█████████▋| 608/624 [00:07<00:00, 52.90it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:1, acc: 0.469, loss: 1.848\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 1: 100%|██████████| 624/624 [00:07<00:00, 79.55it/s]\n",
            "Epoch 2:   1%|          | 4/624 [00:00<00:15, 38.79it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:2, acc: 0.500, loss: 1.844\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 2:  17%|█▋        | 105/624 [00:02<00:12, 41.53it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:2, acc: 0.688, loss: 1.632\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 2:  33%|███▎      | 208/624 [00:03<00:04, 92.54it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:2, acc: 0.688, loss: 1.638\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 2:  50%|█████     | 315/624 [00:04<00:03, 80.19it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:2, acc: 0.719, loss: 1.541\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 2:  66%|██████▌   | 411/624 [00:05<00:02, 89.48it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:2, acc: 0.719, loss: 1.590\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 2:  82%|████████▏ | 511/624 [00:07<00:01, 89.52it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:2, acc: 0.750, loss: 1.403\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 2:  98%|█████████▊| 610/624 [00:08<00:00, 85.09it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:2, acc: 0.562, loss: 1.495\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 2: 100%|██████████| 624/624 [00:08<00:00, 74.38it/s]\n",
            "Epoch 3:   2%|▏         | 10/624 [00:00<00:06, 97.13it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:3, acc: 0.594, loss: 1.481\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 3:  19%|█▊        | 116/624 [00:01<00:05, 89.76it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:3, acc: 0.781, loss: 1.253\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 3:  34%|███▍      | 212/624 [00:02<00:04, 93.23it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:3, acc: 0.750, loss: 1.312\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 3:  50%|████▉     | 311/624 [00:03<00:03, 86.12it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:3, acc: 0.750, loss: 1.240\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 3:  65%|██████▍   | 403/624 [00:05<00:04, 47.43it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:3, acc: 0.750, loss: 1.345\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 3:  83%|████████▎ | 517/624 [00:08<00:01, 63.73it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:3, acc: 0.844, loss: 1.140\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 3:  98%|█████████▊| 614/624 [00:09<00:00, 81.50it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:3, acc: 0.594, loss: 1.302\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 3: 100%|██████████| 624/624 [00:10<00:00, 62.36it/s]\n",
            "Epoch 4:   2%|▏         | 11/624 [00:00<00:05, 108.37it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:4, acc: 0.625, loss: 1.270\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 4:  19%|█▊        | 116/624 [00:01<00:05, 92.51it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:4, acc: 0.812, loss: 1.058\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 4:  35%|███▌      | 219/624 [00:02<00:04, 94.48it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:4, acc: 0.781, loss: 1.135\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 4:  50%|████▉     | 311/624 [00:03<00:03, 89.77it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:4, acc: 0.781, loss: 1.075\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 4:  67%|██████▋   | 416/624 [00:04<00:02, 99.52it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:4, acc: 0.750, loss: 1.209\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 4:  82%|████████▏ | 512/624 [00:05<00:01, 98.94it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:4, acc: 0.844, loss: 0.988\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 4:  98%|█████████▊| 614/624 [00:06<00:00, 94.98it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:4, acc: 0.625, loss: 1.188\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 4: 100%|██████████| 624/624 [00:06<00:00, 93.91it/s]\n",
            "Epoch 5:   2%|▏         | 10/624 [00:00<00:06, 92.65it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:5, acc: 0.625, loss: 1.144\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 5:  18%|█▊        | 110/624 [00:01<00:05, 93.68it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:5, acc: 0.812, loss: 0.947\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 5:  33%|███▎      | 206/624 [00:03<00:08, 51.05it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:5, acc: 0.781, loss: 1.028\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 5:  51%|█████     | 319/624 [00:05<00:03, 78.41it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:5, acc: 0.812, loss: 0.975\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 5:  67%|██████▋   | 419/624 [00:06<00:02, 94.21it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:5, acc: 0.750, loss: 1.126\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 5:  83%|████████▎ | 519/624 [00:07<00:01, 95.31it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:5, acc: 0.875, loss: 0.891\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 5:  98%|█████████▊| 610/624 [00:08<00:00, 95.70it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:5, acc: 0.656, loss: 1.114\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 5: 100%|██████████| 624/624 [00:08<00:00, 72.69it/s]\n",
            "Epoch 6:   2%|▏         | 11/624 [00:00<00:05, 104.91it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:6, acc: 0.719, loss: 1.061\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 6:  18%|█▊        | 115/624 [00:01<00:05, 97.22it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:6, acc: 0.812, loss: 0.876\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 6:  35%|███▌      | 219/624 [00:02<00:04, 93.97it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:6, acc: 0.781, loss: 0.957\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 6:  50%|████▉     | 310/624 [00:03<00:03, 93.13it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:6, acc: 0.844, loss: 0.909\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 6:  66%|██████▌   | 411/624 [00:04<00:02, 87.45it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:6, acc: 0.750, loss: 1.069\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 6:  82%|████████▏ | 511/624 [00:05<00:01, 85.59it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:6, acc: 0.906, loss: 0.823\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 6:  96%|█████████▌| 600/624 [00:06<00:00, 64.91it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:6, acc: 0.656, loss: 1.061\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 6: 100%|██████████| 624/624 [00:07<00:00, 85.72it/s]\n",
            "Epoch 7:   1%|          | 6/624 [00:00<00:10, 57.43it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:7, acc: 0.688, loss: 1.001\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 7:  17%|█▋        | 108/624 [00:02<00:11, 43.91it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:7, acc: 0.812, loss: 0.826\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 7:  35%|███▍      | 218/624 [00:03<00:04, 89.72it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:7, acc: 0.781, loss: 0.906\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 7:  51%|█████     | 318/624 [00:04<00:03, 93.17it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:7, acc: 0.844, loss: 0.861\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 7:  66%|██████▌   | 413/624 [00:05<00:02, 95.38it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:7, acc: 0.750, loss: 1.029\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 7:  83%|████████▎ | 517/624 [00:06<00:01, 95.44it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:7, acc: 0.906, loss: 0.773\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 7:  98%|█████████▊| 610/624 [00:07<00:00, 93.99it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:7, acc: 0.656, loss: 1.020\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 7: 100%|██████████| 624/624 [00:08<00:00, 77.32it/s]\n",
            "Epoch 8:   2%|▏         | 11/624 [00:00<00:05, 103.59it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:8, acc: 0.656, loss: 0.957\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 8:  19%|█▊        | 116/624 [00:01<00:05, 93.45it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:8, acc: 0.812, loss: 0.790\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 8:  34%|███▍      | 211/624 [00:02<00:04, 92.17it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:8, acc: 0.781, loss: 0.867\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 8:  50%|█████     | 315/624 [00:03<00:03, 97.61it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:8, acc: 0.844, loss: 0.826\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 8:  67%|██████▋   | 418/624 [00:04<00:02, 96.29it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:8, acc: 0.750, loss: 0.998\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 8:  81%|████████  | 505/624 [00:05<00:02, 52.86it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:8, acc: 0.906, loss: 0.734\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 8:  97%|█████████▋| 608/624 [00:08<00:00, 46.25it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:8, acc: 0.688, loss: 0.988\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 8: 100%|██████████| 624/624 [00:08<00:00, 73.35it/s]\n",
            "Epoch 9:   3%|▎         | 20/624 [00:00<00:06, 99.20it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:9, acc: 0.656, loss: 0.922\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 9:  18%|█▊        | 114/624 [00:01<00:05, 97.49it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:9, acc: 0.812, loss: 0.761\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 9:  35%|███▌      | 219/624 [00:02<00:04, 96.58it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:9, acc: 0.812, loss: 0.837\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 9:  50%|█████     | 315/624 [00:03<00:03, 101.49it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:9, acc: 0.844, loss: 0.799\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 9:  67%|██████▋   | 421/624 [00:04<00:02, 101.42it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:9, acc: 0.750, loss: 0.973\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 9:  82%|████████▏ | 514/624 [00:05<00:01, 96.76it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:9, acc: 0.875, loss: 0.703\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch 9:  99%|█████████▉| 618/624 [00:06<00:00, 97.49it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "epoch:9, acc: 0.688, loss: 0.961\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\rEpoch 9: 100%|██████████| 624/624 [00:06<00:00, 95.41it/s]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# 假设你在训练过程中已经累积了准确率和损失的值\n",
        "# accuracies 和 losses 是你已经累积的数据列表\n",
        "\n",
        "# 创建绘图窗口\n",
        "plt.figure(figsize=(12, 6))\n",
        "\n",
        "# 绘制准确率变化图\n",
        "plt.subplot(1, 2, 1)  # 创建1行2列的子图，选择第1个子图\n",
        "plt.plot(accuracies, label='Accuracy', color='#1d6fa5')\n",
        "plt.xlabel('Steps')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.title('Accuracy over Training Steps')\n",
        "plt.legend()\n",
        "\n",
        "# 绘制损失变化图\n",
        "plt.subplot(1, 2, 2)  # 创建1行2列的子图，选择第2个子图\n",
        "plt.plot(losses, label='Loss', color='#e57211')\n",
        "plt.xlabel('Steps')\n",
        "plt.ylabel('Loss')\n",
        "plt.title('Loss over Training Steps')\n",
        "plt.legend()\n",
        "\n",
        "# 显示图像\n",
        "plt.tight_layout()  # 自动调整子图参数以适应图像区域\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 329
        },
        "id": "ET7wOPtDUCYI",
        "outputId": "db0fe089-f177-4fd0-a8bc-4c8a1ddf111c"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1200x600 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "data_test = (test_images.astype(np.float32)-127.5) / 127.5\n",
        "X = data_test\n",
        "y = test_labels\n",
        "\n",
        "output = myNeuralNet.forward(data_test)\n",
        "predictions = np.argmax(output, axis=1)\n",
        "if len(y.shape) == 2:\n",
        "    y = np.argmax(y, axis=1)\n",
        "accuracy = np.mean(predictions==y)\n",
        "print(accuracy)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "C2pR_-QvWMUx",
        "outputId": "10c7613f-04ee-42e3-fb46-1c309aacd30a"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.8166816681668166\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "image = test_images[random.randint(1, 9999)]\n",
        "output = myNeuralNet.forward(image)\n",
        "print(output)\n",
        "prediction = np.argmax(output, axis=1)\n",
        "print(f\"预测结果为{prediction}\")\n",
        "plt.imshow(image.reshape(28, 28))\n",
        "plt.axis('off')\n",
        "plt.show()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 477
        },
        "id": "CyCK1dfmWumK",
        "outputId": "c78321b3-5972-4dca-91c2-2257ef1016d2"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[4.65779226e-060 2.53583658e-154 1.23756685e-047 4.33751753e-106\n",
            "  1.73593534e-082 4.93720114e-078 1.00000000e+000 3.09009650e-166\n",
            "  1.55419199e-064 3.97024364e-116]]\n",
            "预测结果为[6]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAGFCAYAAAASI+9IAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAI40lEQVR4nO3cX6jfdR3H8e/vd85ZnUXD5Z/TKMSmZ7qSWGTO44IgWATelDWmeVEQW5TUohCsi7oNa2bm6MaCwho2NPMixFMQUXNJLonasuCUNqwhEVL+YnQ83+5eNwXb++vvz9lvj8f1efH9sO3w3Ofm02vbtm0AoGma/qQPAMD6IQoAhCgAEKIAQIgCACEKAIQoABCiAEDMnusP7u7vGeU5ABix5bUjZ/0ZNwUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQBCFAAIUQAgRAGAEAUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQBCFAAIUQAgRAGAEAUAYnbSBwBGZ+bSS8ub7xx/pLx59z13lDdbDh4tbxg9NwUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQBCFAAIUQAgRAGAEAUAwiupcJ6Y3fL68ubihwflzeb+fHlz7c0ny5u/HyxPGAM3BQBCFAAIUQAgRAGAEAUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYDwIB6MW6/Xabayf2t58+jl93X6VtWJB7eXNwvN0RGchFfKTQGAEAUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQBCFAAIUQAgPIgHY/bCx27otPvt/vE8bnfVY/vLm233etxuWrgpABCiAECIAgAhCgCEKAAQogBAiAIAIQoAhCgAEKIAQIgCACEKAIQH8eAVGNy8s7y5/cAPRnCS/++Rly4qb6755InyZq28YL1yUwAgRAGAEAUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQBCFAAID+IxnXq98qT/1mvKmzvv+nZ58975QXnTNE2zuLyvvNl+56nyZm1wurxhergpABCiAECIAgAhCgCEKAAQogBAiAIAIQoAhCgAEKIAQIgCACEKAIQoABBeSWUqzS5cVt48+qPvjuAk/+u6X32o027xI0+VN6udvsSFzE0BgBAFAEIUAAhRACBEAYAQBQBCFAAIUQAgRAGAEAUAQhQACFEAIDyIx7p36nM3lje7P/jkCE4yHP/840WddvUn/qDOTQGAEAUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQBCFAAIUQAgPIjH2Ky9c0en3S9vv7u8me9t6PStqudWB+XN1Yf+2ulbq51WUOOmAECIAgAhCgCEKAAQogBAiAIAIQoAhCgAEKIAQIgCACEKAIQoABAexGNsnr1pvtOuy+N2Z9r683HXPvSp8mb7l/9S3qyeera8gXFxUwAgRAGAEAUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQBCFAAIUQAgvJJKJy98fKm8eeCWezt+baa8+PzpG8ubxQPHypv6W6xN84f7r+uwapr+v+q/rq9dqf+/b3bQljcX3/9EecP65KYAQIgCACEKAIQoABCiAECIAgAhCgCEKAAQogBAiAIAIQoAhCgAEB7Eo+lv3Fje3HPHN8qbt2+oP2zX1U8evL68OXP4pfLmh0v1P4dtc8fLm3Faa+oP4n3904vlzY/3vqO8aZqmefl3z3TacW7cFAAIUQAgRAGAEAUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQCi17btOb1+tbu/Z9RnYUJWvrejvPn9u741/INwQbn6px/ttLvytl8P+SQXjuW1I2f9GTcFAEIUAAhRACBEAYAQBQBCFAAIUQAgRAGAEAUAQhQACFEAIEQBgJid9AEYrv6ON5c3j++6r8OXNnbYTJ/3nHxfeXN6+Y2dvrXl2L877aoOP1D/97C5Pz+CkzAJbgoAhCgAEKIAQIgCACEKAIQoABCiAECIAgAhCgCEKAAQogBAiAIA4UG8KfPMZ15d3lw+O32P2y09vbe8ueTW0+XN3OD58uYNq8+VN+P0n7ad9BGYIDcFAEIUAAhRACBEAYAQBQBCFAAIUQAgRAGAEAUAQhQACFEAIEQBgBAFAMIrqetUb25Dp92XbnhoyCeZvOuP31LeXLb3VHnz8mBQ3qx3M1e9qbyZ6/2i/p2e/19OC3+TAIQoABCiAECIAgAhCgCEKAAQogBAiAIAIQoAhCgAEKIAQIgCAOFBvHWqN9Ot1x94zT+GfJLhWXp6b6fdwm1/K2+m8XG72a1XlDe7Hj5R3mzuz5c3jw/myputh9bKG0bPTQGAEAUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQBCFAAIUQAgPIi3TrVt22n359X6Q3BXzG7s9K2qr2w/0mn34YP7hnySybppx2867e7a8v3y5lW9+q/4mXa1vPniFz5R3mw6eqy8YfTcFAAIUQAgRAGAEAUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQCi157jy2u7+3tGfRaGYPD+neXNoa9+rbx5y9yG8obxO/D8Unnz1N1vK282Hfa43flgee3sj1K6KQAQogBAiAIAIQoAhCgAEKIAQIgCACEKAIQoABCiAECIAgAhCgCEB/FoZrZdWd6c/Ozrypt9Sz8rb5qmaRbmXixvfv7iYnmzc9NKefPNlV3lzepjl5Q3TdM0C0/U/xx6J/9U3qwNBuUN5wcP4gFQIgoAhCgAEKIAQIgCACEKAIQoABCiAECIAgAhCgCEKAAQogBAiAIA4ZVUgAuEV1IBKBEFAEIUAAhRACBEAYAQBQBCFAAIUQAgRAGAEAUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQBCFAAIUQAgRAGAEAUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQBCFAAIUQAgRAGAEAUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQBCFAAIUQAgRAGAEAUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQBCFAAIUQAgRAGAEAUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQBCFAAIUQAgRAGAEAUAQhQACFEAIEQBgBAFAEIUAAhRACBEAYAQBQCi17ZtO+lDALA+uCkAEKIAQIgCACEKAIQoABCiAECIAgAhCgCEKAAQ/wUfuuNsFqxS/gAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    }
  ]
}