Skip to content

Instantly share code, notes, and snippets.

@agrhn

agrhn/lenet_tensorflow.ipynb

Created January 19, 2021 12:40
Show Gist options

  • Select an option

    Learn more about clone URLs
  • Save agrhn/c6d2a157ebfc883e462f2d6e2ce2e3ce to your computer and use it in GitHub Desktop.

Select an option

Learn more about clone URLs
Save agrhn/c6d2a157ebfc883e462f2d6e2ce2e3ce to your computer and use it in GitHub Desktop.
Download ZIP
LeNet_TensorFlow.ipynb
Display the source blob
Display the rendered blob
Raw
lenet_tensorflow.ipynb
{
"nbformat": 4,
"nbformat_minor": 0,
"metadata": {
"colab": {
"name": "LeNet_TensorFlow.ipynb",
"provenance": [],
"collapsed_sections": [],
"authorship_tag": "ABX9TyOQvBkXChy6hlJ5+j5qxUV4",
"include_colab_link": true
},
"kernelspec": {
"name": "python3",
"display_name": "Python 3"
},
"accelerator": "GPU"
},
"cells": [
{
"cell_type": "markdown",
"metadata": {
"id": "view-in-github",
"colab_type": "text"
},
"source": [
"<a href=\"https://colab.research.google.com/gist/mrgrhn/c6d2a157ebfc883e462f2d6e2ce2e3ce/lenet_tensorflow.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
]
},
{
"cell_type": "code",
"metadata": {
"id": "MMcWUHT5-eVD"
},
"source": [
"import tensorflow as tf\r\n",
"import matplotlib.pyplot as plt\r\n",
"from tensorflow.keras import datasets, layers, models, losses"
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "fEvJTxr2C55-"
},
"source": [
"(x_train, y_train), (x_test, y_test)=tf.keras.datasets.mnist.load_data()"
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "v5CqAuLuDC94",
"outputId": "c082e3c5-58a4-4577-c6a6-cab6a46406ee"
},
"source": [
"x_train.shape"
],
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"(60000, 28, 28)"
]
},
"metadata": {
"tags": []
},
"execution_count": 3
}
]
},
{
"cell_type": "code",
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "XizrWVlmY1NQ",
"outputId": "79110c67-65e5-46cf-c7ef-12f743e14476"
},
"source": [
"x_train = tf.pad(x_train, [[0, 0], [2,2], [2,2]])/255\r\n",
"x_test = tf.pad(x_test, [[0, 0], [2,2], [2,2]])/255\r\n",
"x_train.shape"
],
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"TensorShape([60000, 32, 32])"
]
},
"metadata": {
"tags": []
},
"execution_count": 4
}
]
},
{
"cell_type": "code",
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "hy_dVa44ZOLk",
"outputId": "27409e22-03ac-46a9-b388-8f832fcdcc42"
},
"source": [
"x_train = tf.expand_dims(x_train, axis=3, name=None)\r\n",
"x_test = tf.expand_dims(x_test, axis=3, name=None)\r\n",
"x_train.shape"
],
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"TensorShape([60000, 32, 32, 1])"
]
},
"metadata": {
"tags": []
},
"execution_count": 5
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "gJuMl9GAnoM5"
},
"source": [
"x_val = x_train[-2000:,:,:,:]\r\n",
"y_val = y_train[-2000:]\r\n",
"x_train = x_train[:-2000,:,:,:]\r\n",
"y_train = y_train[:-2000]"
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "UeaLkOM-XRZA",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "38661b63-7ded-4ac8-ce7f-a8c007fa41ee"
},
"source": [
"model = models.Sequential()\r\n",
"model.add(layers.Conv2D(6, 5, activation='tanh', input_shape=x_train.shape[1:]))\r\n",
"model.add(layers.AveragePooling2D(2))\r\n",
"model.add(layers.Activation('sigmoid'))\r\n",
"model.add(layers.Conv2D(16, 5, activation='tanh'))\r\n",
"model.add(layers.AveragePooling2D(2))\r\n",
"model.add(layers.Activation('sigmoid'))\r\n",
"model.add(layers.Conv2D(120, 5, activation='tanh'))\r\n",
"model.add(layers.Flatten())\r\n",
"model.add(layers.Dense(84, activation='tanh'))\r\n",
"model.add(layers.Dense(10, activation='softmax'))\r\n",
"model.summary()"
],
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"text": [
"Model: \"sequential\"\n",
"_________________________________________________________________\n",
"Layer (type) Output Shape Param # \n",
"=================================================================\n",
"conv2d (Conv2D) (None, 28, 28, 6) 156 \n",
"_________________________________________________________________\n",
"average_pooling2d (AveragePo (None, 14, 14, 6) 0 \n",
"_________________________________________________________________\n",
"activation (Activation) (None, 14, 14, 6) 0 \n",
"_________________________________________________________________\n",
"conv2d_1 (Conv2D) (None, 10, 10, 16) 2416 \n",
"_________________________________________________________________\n",
"average_pooling2d_1 (Average (None, 5, 5, 16) 0 \n",
"_________________________________________________________________\n",
"activation_1 (Activation) (None, 5, 5, 16) 0 \n",
"_________________________________________________________________\n",
"conv2d_2 (Conv2D) (None, 1, 1, 120) 48120 \n",
"_________________________________________________________________\n",
"flatten (Flatten) (None, 120) 0 \n",
"_________________________________________________________________\n",
"dense (Dense) (None, 84) 10164 \n",
"_________________________________________________________________\n",
"dense_1 (Dense) (None, 10) 850 \n",
"=================================================================\n",
"Total params: 61,706\n",
"Trainable params: 61,706\n",
"Non-trainable params: 0\n",
"_________________________________________________________________\n"
],
"name": "stdout"
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "pOAJW5ByhYCl"
},
"source": [
"model.compile(optimizer='adam', loss=losses.sparse_categorical_crossentropy, metrics=['accuracy'])"
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "Xv-ml2R6hw_V",
"outputId": "216079d9-5316-457b-fff7-07052aa09efe"
},
"source": [
"history = model.fit(x_train, y_train, batch_size=64, epochs=40, validation_data=(x_val, y_val))"
],
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"text": [
"Epoch 1/40\n",
"907/907 [==============================] - 11s 6ms/step - loss: 1.8860 - accuracy: 0.2950 - val_loss: 0.2266 - val_accuracy: 0.9450\n",
"Epoch 2/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.3354 - accuracy: 0.8943 - val_loss: 0.1769 - val_accuracy: 0.9490\n",
"Epoch 3/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.2478 - accuracy: 0.9208 - val_loss: 0.1392 - val_accuracy: 0.9605\n",
"Epoch 4/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.2073 - accuracy: 0.9335 - val_loss: 0.1354 - val_accuracy: 0.9655\n",
"Epoch 5/40\n",
"907/907 [==============================] - 5s 5ms/step - loss: 0.1788 - accuracy: 0.9443 - val_loss: 0.0892 - val_accuracy: 0.9790\n",
"Epoch 6/40\n",
"907/907 [==============================] - 5s 5ms/step - loss: 0.1574 - accuracy: 0.9502 - val_loss: 0.0987 - val_accuracy: 0.9725\n",
"Epoch 7/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.1406 - accuracy: 0.9550 - val_loss: 0.1154 - val_accuracy: 0.9690\n",
"Epoch 8/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.1277 - accuracy: 0.9603 - val_loss: 0.0865 - val_accuracy: 0.9800\n",
"Epoch 9/40\n",
"907/907 [==============================] - 5s 5ms/step - loss: 0.1116 - accuracy: 0.9653 - val_loss: 0.0886 - val_accuracy: 0.9795\n",
"Epoch 10/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.1180 - accuracy: 0.9629 - val_loss: 0.0950 - val_accuracy: 0.9760\n",
"Epoch 11/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.1028 - accuracy: 0.9675 - val_loss: 0.0814 - val_accuracy: 0.9820\n",
"Epoch 12/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0948 - accuracy: 0.9708 - val_loss: 0.0719 - val_accuracy: 0.9820\n",
"Epoch 13/40\n",
"907/907 [==============================] - 5s 5ms/step - loss: 0.0886 - accuracy: 0.9718 - val_loss: 0.0655 - val_accuracy: 0.9860\n",
"Epoch 14/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0876 - accuracy: 0.9720 - val_loss: 0.0786 - val_accuracy: 0.9805\n",
"Epoch 15/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0807 - accuracy: 0.9739 - val_loss: 0.0710 - val_accuracy: 0.9870\n",
"Epoch 16/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0844 - accuracy: 0.9729 - val_loss: 0.0628 - val_accuracy: 0.9880\n",
"Epoch 17/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0729 - accuracy: 0.9758 - val_loss: 0.0561 - val_accuracy: 0.9875\n",
"Epoch 18/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0699 - accuracy: 0.9779 - val_loss: 0.0750 - val_accuracy: 0.9825\n",
"Epoch 19/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0714 - accuracy: 0.9768 - val_loss: 0.0746 - val_accuracy: 0.9840\n",
"Epoch 20/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0765 - accuracy: 0.9750 - val_loss: 0.0575 - val_accuracy: 0.9855\n",
"Epoch 21/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0679 - accuracy: 0.9774 - val_loss: 0.0775 - val_accuracy: 0.9830\n",
"Epoch 22/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0653 - accuracy: 0.9783 - val_loss: 0.0532 - val_accuracy: 0.9895\n",
"Epoch 23/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0638 - accuracy: 0.9787 - val_loss: 0.0625 - val_accuracy: 0.9885\n",
"Epoch 24/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0606 - accuracy: 0.9796 - val_loss: 0.0606 - val_accuracy: 0.9865\n",
"Epoch 25/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0590 - accuracy: 0.9805 - val_loss: 0.0761 - val_accuracy: 0.9790\n",
"Epoch 26/40\n",
"907/907 [==============================] - 5s 5ms/step - loss: 0.0656 - accuracy: 0.9785 - val_loss: 0.0620 - val_accuracy: 0.9850\n",
"Epoch 27/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0520 - accuracy: 0.9828 - val_loss: 0.0633 - val_accuracy: 0.9835\n",
"Epoch 28/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0546 - accuracy: 0.9823 - val_loss: 0.0738 - val_accuracy: 0.9825\n",
"Epoch 29/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0560 - accuracy: 0.9815 - val_loss: 0.0555 - val_accuracy: 0.9895\n",
"Epoch 30/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0496 - accuracy: 0.9842 - val_loss: 0.0536 - val_accuracy: 0.9880\n",
"Epoch 31/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0503 - accuracy: 0.9839 - val_loss: 0.0716 - val_accuracy: 0.9835\n",
"Epoch 32/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0475 - accuracy: 0.9843 - val_loss: 0.0559 - val_accuracy: 0.9880\n",
"Epoch 33/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0477 - accuracy: 0.9843 - val_loss: 0.0472 - val_accuracy: 0.9900\n",
"Epoch 34/40\n",
"907/907 [==============================] - 5s 5ms/step - loss: 0.0471 - accuracy: 0.9847 - val_loss: 0.0600 - val_accuracy: 0.9880\n",
"Epoch 35/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0425 - accuracy: 0.9859 - val_loss: 0.0726 - val_accuracy: 0.9850\n",
"Epoch 36/40\n",
"907/907 [==============================] - 5s 5ms/step - loss: 0.0437 - accuracy: 0.9858 - val_loss: 0.0472 - val_accuracy: 0.9900\n",
"Epoch 37/40\n",
"907/907 [==============================] - 5s 5ms/step - loss: 0.0411 - accuracy: 0.9860 - val_loss: 0.0511 - val_accuracy: 0.9885\n",
"Epoch 38/40\n",
"907/907 [==============================] - 4s 5ms/step - loss: 0.0399 - accuracy: 0.9867 - val_loss: 0.0530 - val_accuracy: 0.9885\n",
"Epoch 39/40\n",
"907/907 [==============================] - 5s 5ms/step - loss: 0.0381 - accuracy: 0.9869 - val_loss: 0.0523 - val_accuracy: 0.9900\n",
"Epoch 40/40\n",
"907/907 [==============================] - 5s 5ms/step - loss: 0.0375 - accuracy: 0.9875 - val_loss: 0.0428 - val_accuracy: 0.9915\n"
],
"name": "stdout"
}
]
},
{
"cell_type": "code",
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 896
},
"id": "AfX0d7zyocFw",
"outputId": "b92f5a88-5045-4941-9616-a812f6a7480f"
},
"source": [
"fig, axs = plt.subplots(2, 1, figsize=(15,15))\r\n",
"\r\n",
"axs[0].plot(history.history['loss'])\r\n",
"axs[0].plot(history.history['val_loss'])\r\n",
"axs[0].title.set_text('Training Loss vs Validation Loss')\r\n",
"axs[0].legend(['Train', 'Val'])\r\n",
"\r\n",
"axs[1].plot(history.history['accuracy'])\r\n",
"axs[1].plot(history.history['val_accuracy'])\r\n",
"axs[1].title.set_text('Training Accuracy vs Validation Accuracy')\r\n",
"axs[1].legend(['Train', 'Val'])\r\n"
],
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"<matplotlib.legend.Legend at 0x7f51d0179dd8>"
]
},
"metadata": {
"tags": []
},
"execution_count": 10
},
{
"output_type": "display_data",
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 1080x1080 with 2 Axes>"
]
},
"metadata": {
"tags": [],
"needs_background": "light"
}
}
]
},
{
"cell_type": "code",
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "iLQqDljamZTV",
"outputId": "4d65a32a-5fec-4b85-8aab-5a26fedd9166"
},
"source": [
"model.evaluate(x_test, y_test)"
],
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"text": [
"313/313 [==============================] - 1s 3ms/step - loss: 0.0549 - accuracy: 0.9825\n"
],
"name": "stdout"
},
{
"output_type": "execute_result",
"data": {
"text/plain": [
"[0.05486104264855385, 0.9825000166893005]"
]
},
"metadata": {
"tags": []
},
"execution_count": 11
}
]
}
]
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment