wehub-resource-sync
319 lines
11 KiB
Plaintext
319 lines
11 KiB
Plaintext
{
|
|
"cells": [
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {},
|
|
"source": [
|
|
"Deep Learning Models -- A collection of various deep learning architectures, models, and tips for TensorFlow and PyTorch in Jupyter Notebooks.\n",
|
|
"- Author: Sebastian Raschka\n",
|
|
"- GitHub Repository: https://github.com/rasbt/deeplearning-models"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 1,
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Sebastian Raschka \n",
|
|
"\n",
|
|
"CPython 3.6.0\n",
|
|
"IPython 6.0.0\n",
|
|
"\n",
|
|
"tensorflow 1.1.0\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"%load_ext watermark\n",
|
|
"%watermark -a 'Sebastian Raschka' -v -p tensorflow"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {},
|
|
"source": [
|
|
"# Model Zoo -- Multilayer Perceptron"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {},
|
|
"source": [
|
|
"### Low-level Implementation"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 2,
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Extracting ./train-images-idx3-ubyte.gz\n",
|
|
"Extracting ./train-labels-idx1-ubyte.gz\n",
|
|
"Extracting ./t10k-images-idx3-ubyte.gz\n",
|
|
"Extracting ./t10k-labels-idx1-ubyte.gz\n",
|
|
"Epoch: 001 | AvgCost: 0.349 | Train/Valid ACC: 0.945/0.944\n",
|
|
"Epoch: 002 | AvgCost: 0.164 | Train/Valid ACC: 0.962/0.961\n",
|
|
"Epoch: 003 | AvgCost: 0.118 | Train/Valid ACC: 0.973/0.969\n",
|
|
"Epoch: 004 | AvgCost: 0.092 | Train/Valid ACC: 0.979/0.971\n",
|
|
"Epoch: 005 | AvgCost: 0.075 | Train/Valid ACC: 0.983/0.974\n",
|
|
"Epoch: 006 | AvgCost: 0.061 | Train/Valid ACC: 0.985/0.976\n",
|
|
"Epoch: 007 | AvgCost: 0.052 | Train/Valid ACC: 0.988/0.976\n",
|
|
"Epoch: 008 | AvgCost: 0.043 | Train/Valid ACC: 0.991/0.978\n",
|
|
"Epoch: 009 | AvgCost: 0.037 | Train/Valid ACC: 0.993/0.980\n",
|
|
"Epoch: 010 | AvgCost: 0.030 | Train/Valid ACC: 0.994/0.979\n",
|
|
"Test ACC: 0.975\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"import tensorflow as tf\n",
|
|
"from tensorflow.examples.tutorials.mnist import input_data\n",
|
|
"\n",
|
|
"\n",
|
|
"##########################\n",
|
|
"### DATASET\n",
|
|
"##########################\n",
|
|
"\n",
|
|
"mnist = input_data.read_data_sets(\"./\", one_hot=True)\n",
|
|
"\n",
|
|
"\n",
|
|
"##########################\n",
|
|
"### SETTINGS\n",
|
|
"##########################\n",
|
|
"\n",
|
|
"# Hyperparameters\n",
|
|
"learning_rate = 0.1\n",
|
|
"training_epochs = 10\n",
|
|
"batch_size = 64\n",
|
|
"\n",
|
|
"# Architecture\n",
|
|
"n_hidden_1 = 128\n",
|
|
"n_hidden_2 = 256\n",
|
|
"n_input = 784\n",
|
|
"n_classes = 10\n",
|
|
"\n",
|
|
"\n",
|
|
"##########################\n",
|
|
"### GRAPH DEFINITION\n",
|
|
"##########################\n",
|
|
"\n",
|
|
"g = tf.Graph()\n",
|
|
"with g.as_default():\n",
|
|
"\n",
|
|
" # Input data\n",
|
|
" tf_x = tf.placeholder(tf.float32, [None, n_input], name='features')\n",
|
|
" tf_y = tf.placeholder(tf.float32, [None, n_classes], name='targets')\n",
|
|
"\n",
|
|
" # Model parameters\n",
|
|
" weights = {\n",
|
|
" 'h1': tf.Variable(tf.truncated_normal([n_input, n_hidden_1], stddev=0.1)),\n",
|
|
" 'h2': tf.Variable(tf.truncated_normal([n_hidden_1, n_hidden_2], stddev=0.1)),\n",
|
|
" 'out': tf.Variable(tf.truncated_normal([n_hidden_2, n_classes], stddev=0.1))\n",
|
|
" }\n",
|
|
" biases = {\n",
|
|
" 'b1': tf.Variable(tf.zeros([n_hidden_1])),\n",
|
|
" 'b2': tf.Variable(tf.zeros([n_hidden_2])),\n",
|
|
" 'out': tf.Variable(tf.zeros([n_classes]))\n",
|
|
" }\n",
|
|
"\n",
|
|
" # Multilayer perceptron\n",
|
|
" layer_1 = tf.add(tf.matmul(tf_x, weights['h1']), biases['b1'])\n",
|
|
" layer_1 = tf.nn.relu(layer_1)\n",
|
|
" layer_2 = tf.add(tf.matmul(layer_1, weights['h2']), biases['b2'])\n",
|
|
" layer_2 = tf.nn.relu(layer_2)\n",
|
|
" out_layer = tf.matmul(layer_2, weights['out']) + biases['out']\n",
|
|
"\n",
|
|
" # Loss and optimizer\n",
|
|
" loss = tf.nn.softmax_cross_entropy_with_logits(logits=out_layer, labels=tf_y)\n",
|
|
" cost = tf.reduce_mean(loss, name='cost')\n",
|
|
" optimizer = tf.train.GradientDescentOptimizer(learning_rate=learning_rate)\n",
|
|
" train = optimizer.minimize(cost, name='train')\n",
|
|
"\n",
|
|
" # Prediction\n",
|
|
" correct_prediction = tf.equal(tf.argmax(tf_y, 1), tf.argmax(out_layer, 1))\n",
|
|
" accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32), name='accuracy')\n",
|
|
"\n",
|
|
" \n",
|
|
"##########################\n",
|
|
"### TRAINING & EVALUATION\n",
|
|
"##########################\n",
|
|
"\n",
|
|
"with tf.Session(graph=g) as sess:\n",
|
|
" sess.run(tf.global_variables_initializer())\n",
|
|
"\n",
|
|
" for epoch in range(training_epochs):\n",
|
|
" avg_cost = 0.\n",
|
|
" total_batch = mnist.train.num_examples // batch_size\n",
|
|
"\n",
|
|
" for i in range(total_batch):\n",
|
|
" batch_x, batch_y = mnist.train.next_batch(batch_size)\n",
|
|
" _, c = sess.run(['train', 'cost:0'], feed_dict={'features:0': batch_x,\n",
|
|
" 'targets:0': batch_y})\n",
|
|
" avg_cost += c\n",
|
|
" \n",
|
|
" train_acc = sess.run('accuracy:0', feed_dict={'features:0': mnist.train.images,\n",
|
|
" 'targets:0': mnist.train.labels})\n",
|
|
" valid_acc = sess.run('accuracy:0', feed_dict={'features:0': mnist.validation.images,\n",
|
|
" 'targets:0': mnist.validation.labels}) \n",
|
|
" \n",
|
|
" print(\"Epoch: %03d | AvgCost: %.3f\" % (epoch + 1, avg_cost / (i + 1)), end=\"\")\n",
|
|
" print(\" | Train/Valid ACC: %.3f/%.3f\" % (train_acc, valid_acc))\n",
|
|
" \n",
|
|
" test_acc = sess.run(accuracy, feed_dict={'features:0': mnist.test.images,\n",
|
|
" 'targets:0': mnist.test.labels})\n",
|
|
" print('Test ACC: %.3f' % test_acc)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {},
|
|
"source": [
|
|
"### tensorflow.layers Abstraction"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 3,
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Extracting ./train-images-idx3-ubyte.gz\n",
|
|
"Extracting ./train-labels-idx1-ubyte.gz\n",
|
|
"Extracting ./t10k-images-idx3-ubyte.gz\n",
|
|
"Extracting ./t10k-labels-idx1-ubyte.gz\n",
|
|
"Epoch: 001 | AvgCost: 0.344 | Train/Valid ACC: 0.946/0.946\n",
|
|
"Epoch: 002 | AvgCost: 0.159 | Train/Valid ACC: 0.965/0.965\n",
|
|
"Epoch: 003 | AvgCost: 0.115 | Train/Valid ACC: 0.973/0.969\n",
|
|
"Epoch: 004 | AvgCost: 0.090 | Train/Valid ACC: 0.979/0.973\n",
|
|
"Epoch: 005 | AvgCost: 0.073 | Train/Valid ACC: 0.978/0.971\n",
|
|
"Epoch: 006 | AvgCost: 0.062 | Train/Valid ACC: 0.985/0.975\n",
|
|
"Epoch: 007 | AvgCost: 0.051 | Train/Valid ACC: 0.990/0.977\n",
|
|
"Epoch: 008 | AvgCost: 0.043 | Train/Valid ACC: 0.992/0.979\n",
|
|
"Epoch: 009 | AvgCost: 0.036 | Train/Valid ACC: 0.993/0.978\n",
|
|
"Epoch: 010 | AvgCost: 0.030 | Train/Valid ACC: 0.991/0.975\n",
|
|
"Test ACC: 0.975\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"import tensorflow as tf\n",
|
|
"from tensorflow.examples.tutorials.mnist import input_data\n",
|
|
"\n",
|
|
"\n",
|
|
"##########################\n",
|
|
"### DATASET\n",
|
|
"##########################\n",
|
|
"\n",
|
|
"mnist = input_data.read_data_sets(\"./\", one_hot=True)\n",
|
|
"\n",
|
|
"\n",
|
|
"##########################\n",
|
|
"### SETTINGS\n",
|
|
"##########################\n",
|
|
"\n",
|
|
"# Hyperparameters\n",
|
|
"learning_rate = 0.1\n",
|
|
"training_epochs = 10\n",
|
|
"batch_size = 64\n",
|
|
"\n",
|
|
"# Architecture\n",
|
|
"n_hidden_1 = 128\n",
|
|
"n_hidden_2 = 256\n",
|
|
"n_input = 784\n",
|
|
"n_classes = 10\n",
|
|
"\n",
|
|
"\n",
|
|
"##########################\n",
|
|
"### GRAPH DEFINITION\n",
|
|
"##########################\n",
|
|
"\n",
|
|
"g = tf.Graph()\n",
|
|
"with g.as_default():\n",
|
|
"\n",
|
|
" # Input data\n",
|
|
" tf_x = tf.placeholder(tf.float32, [None, n_input], name='features')\n",
|
|
" tf_y = tf.placeholder(tf.float32, [None, n_classes], name='targets')\n",
|
|
"\n",
|
|
" # Multilayer perceptron\n",
|
|
" layer_1 = tf.layers.dense(tf_x, n_hidden_1, activation=tf.nn.relu, \n",
|
|
" kernel_initializer=tf.truncated_normal_initializer(stddev=0.1))\n",
|
|
" layer_2 = tf.layers.dense(layer_1, n_hidden_2, activation=tf.nn.relu,\n",
|
|
" kernel_initializer=tf.truncated_normal_initializer(stddev=0.1))\n",
|
|
" out_layer = tf.layers.dense(layer_2, n_classes, activation=None)\n",
|
|
"\n",
|
|
" # Loss and optimizer\n",
|
|
" loss = tf.nn.softmax_cross_entropy_with_logits(logits=out_layer, labels=tf_y)\n",
|
|
" cost = tf.reduce_mean(loss, name='cost')\n",
|
|
" optimizer = tf.train.GradientDescentOptimizer(learning_rate=learning_rate)\n",
|
|
" train = optimizer.minimize(cost, name='train')\n",
|
|
"\n",
|
|
" # Prediction\n",
|
|
" correct_prediction = tf.equal(tf.argmax(tf_y, 1), tf.argmax(out_layer, 1))\n",
|
|
" accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32), name='accuracy')\n",
|
|
"\n",
|
|
"\n",
|
|
"##########################\n",
|
|
"### TRAINING & EVALUATION\n",
|
|
"##########################\n",
|
|
" \n",
|
|
"with tf.Session(graph=g) as sess:\n",
|
|
" sess.run(tf.global_variables_initializer())\n",
|
|
"\n",
|
|
" for epoch in range(training_epochs):\n",
|
|
" avg_cost = 0.\n",
|
|
" total_batch = mnist.train.num_examples // batch_size\n",
|
|
"\n",
|
|
" for i in range(total_batch):\n",
|
|
" batch_x, batch_y = mnist.train.next_batch(batch_size)\n",
|
|
" _, c = sess.run(['train', 'cost:0'], feed_dict={'features:0': batch_x,\n",
|
|
" 'targets:0': batch_y})\n",
|
|
" avg_cost += c\n",
|
|
" \n",
|
|
" train_acc = sess.run('accuracy:0', feed_dict={'features:0': mnist.train.images,\n",
|
|
" 'targets:0': mnist.train.labels})\n",
|
|
" valid_acc = sess.run('accuracy:0', feed_dict={'features:0': mnist.validation.images,\n",
|
|
" 'targets:0': mnist.validation.labels}) \n",
|
|
" \n",
|
|
" print(\"Epoch: %03d | AvgCost: %.3f\" % (epoch + 1, avg_cost / (i + 1)), end=\"\")\n",
|
|
" print(\" | Train/Valid ACC: %.3f/%.3f\" % (train_acc, valid_acc))\n",
|
|
" \n",
|
|
" test_acc = sess.run('accuracy:0', feed_dict={'features:0': mnist.test.images,\n",
|
|
" 'targets:0': mnist.test.labels})\n",
|
|
" print('Test ACC: %.3f' % test_acc)"
|
|
]
|
|
}
|
|
],
|
|
"metadata": {
|
|
"kernelspec": {
|
|
"display_name": "Python 3",
|
|
"language": "python",
|
|
"name": "python3"
|
|
},
|
|
"language_info": {
|
|
"codemirror_mode": {
|
|
"name": "ipython",
|
|
"version": 3
|
|
},
|
|
"file_extension": ".py",
|
|
"mimetype": "text/x-python",
|
|
"name": "python",
|
|
"nbconvert_exporter": "python",
|
|
"pygments_lexer": "ipython3",
|
|
"version": "3.7.1"
|
|
}
|
|
},
|
|
"nbformat": 4,
|
|
"nbformat_minor": 2
|
|
}
|