Skip to content

Instantly share code, notes, and snippets.

@Breta01
Last active October 22, 2020 09:02
Show Gist options
  • Save Breta01/f205a9d27090c18d394fbaab98de7c65 to your computer and use it in GitHub Desktop.
Save Breta01/f205a9d27090c18d394fbaab98de7c65 to your computer and use it in GitHub Desktop.
Example of importing multiple TensorFlow modules
Display the source blob
Display the rendered blob
Raw
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Importing Multiple Modules\n",
"## First we create and train one of the models"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"import tensorflow as tf\n",
"import random\n",
"import matplotlib.pylab as plt"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Generate random data for training"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAD8CAYAAAB0IB+mAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/j9GIw9SaiKY/gAAAABJRU5ErkJggg==\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f73e2f0e278>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"data = [0]\n",
"expected = [0]\n",
"for i in range(100):\n",
" expected.append(expected[-1] + random.randint(-10, 10))\n",
" data.append(i)\n",
" \n",
"plt.plot(data, expected)\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create the graph"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"### Linear Regression ###\n",
"# Imput placeholders\n",
"x = tf.placeholder(tf.float32, name='x')\n",
"y = tf.placeholder(tf.float32, name='y')\n",
"# Model parameters\n",
"W1 = tf.Variable([0.1], tf.float32)\n",
"W2 = tf.Variable([0.1], tf.float32)\n",
"W3 = tf.Variable([0.1], tf.float32)\n",
"b = tf.Variable([0.1], tf.float32)\n",
"\n",
"# Output\n",
"linear_model = tf.identity(W1 * x + W2 * x**2 + W3 * x**3 + b,\n",
" name='activation_opt')\n",
"\n",
"# Loss\n",
"loss = tf.reduce_sum(tf.square(linear_model - y), name='loss')\n",
"# Optimizer and training step\n",
"optimizer = tf.train.AdamOptimizer(0.001)\n",
"train = optimizer.minimize(loss, name='train_step')\n",
"\n",
"# Remember output operation for later aplication\n",
"# Adding it to a collections for easy acces\n",
"# This is not required if you NAME your output operation\n",
"tf.add_to_collection(\"activation\", linear_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Train + Save"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAD8CAYAAAB0IB+mAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhAK+iEhE/D9E8nvrWOn3yAAAAABJRU5ErkJggg==\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f73e2f36390>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Start the session\n",
"init = tf.global_variables_initializer()\n",
"sess = tf.Session()\n",
"sess.run(init)\n",
"# Create saver\n",
"saver = tf.train.Saver()\n",
"\n",
"# Training loop\n",
"for i in range(10000):\n",
" sess.run(train, {x: data, y: expected})\n",
" if i % 1000 == 0:\n",
" # You can also save checkpoints using global_step variable\n",
" saver.save(sess, \"models/model_name\", global_step=i)\n",
"\n",
"# Save TensorFlow graph into path models/model_name\n",
"saver.save(sess, \"models/model_name\")\n",
"\n",
"\n",
"predicted = sess.run(linear_model, {x: [i for i in range(100)]})\n",
"\n",
"plt.plot(data, expected, 'o', markersize=2)\n",
"plt.plot(predicted, 'r')\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Session doesn't exist\n"
]
}
],
"source": [
"# Close previous session and reset graph\n",
"tf.reset_default_graph()\n",
"sess.close()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Importing (in different file)"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"class ImportGraph():\n",
" \"\"\" Importing and running isolated TF graph \"\"\"\n",
" def __init__(self, loc):\n",
" # Create local graph and use it in the session\n",
" self.graph = tf.Graph()\n",
" self.sess = tf.Session(graph=self.graph)\n",
" with self.graph.as_default():\n",
" # Import saved model from location 'loc' into local graph\n",
" saver = tf.train.import_meta_graph(loc + '.meta',\n",
" clear_devices=True)\n",
" saver.restore(self.sess, loc)\n",
" # There are TWO options how to get activation operation:\n",
" # FROM SAVED COLLECTION: \n",
" self.activation = tf.get_collection('activation')[0]\n",
" # BY NAME:\n",
" self.activation = self.graph.get_operation_by_name('activation_opt').outputs[0]\n",
"\n",
" def run(self, data):\n",
" \"\"\" Running the activation operation previously imported \"\"\"\n",
" return self.sess.run(self.activation, feed_dict={\"x:0\": data})"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAD8CAYAAAB0IB+mAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJTOgAAAAAElFTkSuQmCC\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f73e03df588>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"model_1 = ImportGraph('models/model_name')\n",
"model_2 = ImportGraph('models/different_model')\n",
"\n",
"# Application of two different models, ploting results\n",
"data = range(101)\n",
"plt.plot(model_1.run(data), 'r')\n",
"plt.plot(model_2.run(data), 'g')\n",
"plt.plot(data, expected, 'o', markersize=2)\n",
"plt.show()"
]
}
],
"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": "3.6.4"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment