Navigation Menu

Skip to content

Instantly share code, notes, and snippets.

@frh-1

frh-1/gist:673988e2c6b27b4ca9438142f24f9511

Created March 9, 2017 19:00
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save frh-1/673988e2c6b27b4ca9438142f24f9511 to your computer and use it in GitHub Desktop.
Save frh-1/673988e2c6b27b4ca9438142f24f9511 to your computer and use it in GitHub Desktop.
Download ZIP
State Farm vgg16 adapted model
Raw
gistfile1.txt
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"u'/home/ubuntu/courses/deeplearning1/nbs/lesson1/statefarm'"
]
},
"execution_count": 1,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"%pwd"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"import os, sys\n",
"\n",
"HOME_DIR = os.getcwd()"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"Using gpu device 0: Tesla K80 (CNMeM is disabled, cuDNN 5103)\n",
"/home/ubuntu/anaconda2/lib/python2.7/site-packages/theano/sandbox/cuda/__init__.py:600: UserWarning: Your cuDNN version is more recent than the one Theano officially supports. If you see any problems, try updating Theano or downgrading cuDNN to version 5.\n",
" warnings.warn(warn)\n",
"Using Theano backend.\n"
]
}
],
"source": [
"# allow relative imports to directories above statefarm\n",
"sys.path.insert(1, os.path.join(sys.path[0], '../..'))\n",
"\n",
"from utils import *\n",
"from vgg16 import Vgg16"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"driver_imgs_list.csv \u001b[0m\u001b[01;34msample\u001b[0m/ statefarm.ipynb \u001b[01;34mtrain\u001b[0m/\r\n",
"\u001b[01;34mresults\u001b[0m/ sample_submission.csv \u001b[01;34mtest\u001b[0m/ \u001b[01;34mvalid\u001b[0m/\r\n"
]
}
],
"source": [
"%ls"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## PREPARE THE DATA\n",
"Creat the appropriate training, validation and sample directories.\n",
"Move the image files into the correct directories."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## FINE TUNING AND TRAINING VGG MODEL"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"/home/ubuntu/courses/deeplearning1/nbs/lesson1/statefarm\n"
]
}
],
"source": [
"%cd $HOME_DIR\n",
"\n",
"# Set directory variables\n",
"path = HOME_DIR + '/'#sample/'\n",
"train_path = path + 'train/'\n",
"valid_path = path + 'valid/'\n",
"test_path = HOME_DIR +'/test/'\n",
"results_path = HOME_DIR + '/results/'"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"model_path = HOME_DIR + '/models/'\n",
"if not os.path.exists(model_path): os.mkdir(model_path)"
]
},
{
"cell_type": "code",
"execution_count": 27,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"# Import vgg helper class\n",
"vgg = Vgg16()"
]
},
{
"cell_type": "code",
"execution_count": 30,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"# Set constants\n",
"batch_size = 64\n",
"no_of_epochs = 3"
]
},
{
"cell_type": "code",
"execution_count": 29,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Found 19977 images belonging to 10 classes.\n",
"Found 2447 images belonging to 10 classes.\n"
]
}
],
"source": [
"# Finetune the model\n",
"# The get_batches function gets batches plus returns the number of classes\n",
"# The .finetune method pops the final dense layer in the VGG model and replace with a new dense layer\n",
"# with a number of units equal to the number of batches\n",
"\n",
"train_batches = vgg.get_batches(train_path, batch_size = batch_size)\n",
"valid_batches = vgg.get_batches(valid_path, batch_size=batch_size*2)\n",
"vgg.finetune(train_batches)\n",
"\n",
"vgg.model.optimizer.lr = 0.01"
]
},
{
"cell_type": "code",
"execution_count": 31,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Running epoch no. 0\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 563s - loss: 10.5082 - acc: 0.2937 - val_loss: 10.5327 - val_acc: 0.3036\n",
"Running epoch no. 1\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 565s - loss: 10.5092 - acc: 0.3237 - val_loss: 10.9144 - val_acc: 0.2889\n",
"Running epoch no. 2\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 564s - loss: 10.5387 - acc: 0.3283 - val_loss: 11.1385 - val_acc: 0.2881\n",
"Completed 3 fit operations\n"
]
}
],
"source": [
"# Fit the model and test performance against the validation set\n",
"latest_weights_filename = None\n",
"for epoch in range(no_of_epochs):\n",
" print \"Running epoch no. %d\" %epoch\n",
" vgg.fit(train_batches, valid_batches, nb_epoch=1)\n",
" latest_weights_filename = 'ft_n_%d.h5' %epoch\n",
" vgg.model.save_weights(results_path + latest_weights_filename)\n",
"print \"Completed %s fit operations\" %no_of_epochs"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Split off convolution layer and calculate features for efficiency"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"# Get the last convolution layer\n",
"layers = vgg.model.layers\n",
"last_conv_idx = [index for index, layer in enumerate(layers) if type(layer) is Convolution2D][-1]"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"# Separate out the convolution block and create a convolution only model\n",
"conv_layers = vgg.model.layers[:last_conv_idx+1]\n",
"conv_model = Sequential(conv_layers)"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Found 19977 images belonging to 10 classes.\n",
"Found 2447 images belonging to 10 classes.\n"
]
}
],
"source": [
"batch_size = 64\n",
"train_batches = get_batches(train_path, batch_size = batch_size,shuffle=False)\n",
"val_batches = get_batches(valid_path, batch_size=batch_size, shuffle=False)\n",
"\n",
"train_labels = train_batches.classes\n",
"val_labels = val_batches.classes\n",
"train_classes = onehot(train_labels)\n",
"val_classes = onehot(val_labels)"
]
},
{
"cell_type": "code",
"execution_count": 35,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"train_features = conv_model.predict_generator(train_batches, train_batches.nb_sample)\n",
"val_features = conv_model.predict_generator(val_batches, val_batches.nb_sample)"
]
},
{
"cell_type": "code",
"execution_count": 36,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"save_array(model_path+'train_convlayer_features.bc', train_features)\n",
"save_array(model_path+'val_convlayer_features.bc', val_features)\n",
"save_array(model_path+'train_convlayer_labels.bc',train_labels)\n",
"save_array(model_path+'val_convlayer_labels.bc',val_labels)\n",
"save_array(model_path+'train_convlayer_classes.bc',train_classes)\n",
"save_array(model_path+'val_convlayer_classes.bc',train_labels)"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"train_features = load_array(model_path+'train_convlayer_features.bc')\n",
"val_features = load_array(model_path+'val_convlayer_features.bc')\n",
"train_labels = load_array(model_path+'train_convlayer_labels.bc')\n",
"val_labels = load_array(model_path+'val_convlayer_labels.bc')\n",
"train_classes = load_array(model_path+'train_convlayer_classes.bc')\n",
"train_labels = load_array(model_path+'val_convlayer_classes.bc')"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"def get_fc_model(p,lr=0.0001):\n",
" model = Sequential([\n",
" MaxPooling2D(input_shape = conv_layers[-1].output_shape[1:]),\n",
" Flatten(),\n",
" Dense(4096, activation='relu', init='uniform'),\n",
" Dropout(p),\n",
" Dense(4096, activation='relu',init='uniform'),\n",
" Dropout(p),\n",
" Dense(10, activation='softmax',init='uniform')\n",
" ])\n",
" \n",
" for layer in model.layers: layer.trainable = False\n",
" model.layers[-1].trainable = True\n",
" \n",
" model.compile(optimizer=Adam(lr=lr), loss='categorical_crossentropy', metrics=['accuracy'])\n",
" return model"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"fc_model_lst = get_fc_model(p=0.5,lr=0.0001)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Learning Rate\n",
"If the learning rate is too high when the model is still not well fitted, the neurons easily get saturated and then learning ceases - the model gets stuck at the same solution."
]
},
{
"cell_type": "code",
"execution_count": 44,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Running epoch no. 0\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 9s - loss: 14.2683 - acc: 0.1091 - val_loss: 14.5742 - val_acc: 0.0879\n",
"Running epoch no. 1\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 9s - loss: 14.2689 - acc: 0.1093 - val_loss: 14.3676 - val_acc: 0.1058\n",
"Running epoch no. 2\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 8s - loss: 14.1081 - acc: 0.1195 - val_loss: 14.2437 - val_acc: 0.1054\n",
"Running epoch no. 3\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 8s - loss: 14.0560 - acc: 0.1232 - val_loss: 14.5352 - val_acc: 0.0952\n",
"Running epoch no. 4\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 8s - loss: 13.9511 - acc: 0.1293 - val_loss: 14.0530 - val_acc: 0.1177\n",
"Running epoch no. 5\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 8s - loss: 13.8341 - acc: 0.1367 - val_loss: 13.7590 - val_acc: 0.1402\n",
"Running epoch no. 6\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 8s - loss: 13.5647 - acc: 0.1518 - val_loss: 14.0540 - val_acc: 0.1181\n",
"Running epoch no. 7\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 8s - loss: 13.5688 - acc: 0.1526 - val_loss: 13.7386 - val_acc: 0.1434\n",
"Running epoch no. 8\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 8s - loss: 13.3875 - acc: 0.1631 - val_loss: 13.2760 - val_acc: 0.1639\n",
"Running epoch no. 9\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 9s - loss: 13.2454 - acc: 0.1719 - val_loss: 12.9581 - val_acc: 0.1880\n",
"Completed 10 fit operations\n"
]
}
],
"source": [
"# Run more epochs on the fully connected model\n",
"no_of_epochs = 10\n",
"latest_weights_filename = None\n",
"for epoch in range(no_of_epochs):\n",
" print \"Running epoch no. %d\" %epoch\n",
" fc_model_lst.model.fit(train_features,train_classes,batch_size=batch_size,nb_epoch=1,validation_data=(val_features,val_classes))\n",
" latest_weights_filename = 'fc_0_%d.h5' %epoch\n",
" fc_model_lst.model.save_weights(results_path + latest_weights_filename)\n",
"print \"Completed %s fit operations\" %no_of_epochs"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"dense_idx = [index for index, layer in enumerate(fc_layers) if type(layer)==Dense]\n",
"#fc_model.model.layers[dense_idx[-2]].trainable = True"
]
},
{
"cell_type": "code",
"execution_count": 35,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"for layer in fc_model.layers: layer.trainable = False\n",
"fc_model.layers[-1].trainable = True"
]
},
{
"cell_type": "code",
"execution_count": 39,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"fc_model.compile(optimizer=Adam(),loss='categorical_crossentropy', metrics=['accuracy'])"
]
},
{
"cell_type": "code",
"execution_count": 41,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"fc_model.model.optimizer.lr = 0.01"
]
},
{
"cell_type": "code",
"execution_count": 49,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"fc_model.save_weights(results_path+'fc0.h5')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## ADD BATCH NORMALISATION"
]
},
{
"cell_type": "code",
"execution_count": 27,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"opt = RMSprop(lr=0.0001)\n",
"\n",
"def get_bn_model(p):\n",
" model = Sequential([\n",
" MaxPooling2D(input_shape=conv_layers[-1].output_shape[1:]),\n",
" Flatten(),\n",
" Dense(4096,activation='relu'),\n",
" Dropout(p),\n",
" BatchNormalization(),\n",
" Dense(4096,activation='relu'),\n",
" Dropout(p),\n",
" BatchNormalization(),\n",
" Dense(10,activation='softmax')\n",
" ])\n",
" \n",
" model.compile(optimizer=opt,loss='categorical_crossentropy',metrics=['accuracy'])\n",
" return model"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"last_conv_idx=[index for index,layer in enumerate(vgg.model.layers) if type(layer) is Convolution2D][-1]"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"conv_layers = vgg.model.layers[:last_conv_idx+1]"
]
},
{
"cell_type": "code",
"execution_count": 28,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"bn_model=get_bn_model(0.6)"
]
},
{
"cell_type": "code",
"execution_count": 47,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Running epoch no. 0\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 31s - loss: 0.3643 - acc: 0.8858 - val_loss: 0.5988 - val_acc: 0.7912\n",
"Running epoch no. 1\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 31s - loss: 0.0388 - acc: 0.9887 - val_loss: 0.5556 - val_acc: 0.8337\n",
"Running epoch no. 2\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 32s - loss: 0.0222 - acc: 0.9936 - val_loss: 0.6975 - val_acc: 0.7855\n",
"Running epoch no. 3\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 32s - loss: 0.0157 - acc: 0.9957 - val_loss: 0.8910 - val_acc: 0.7912\n",
"Running epoch no. 4\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 32s - loss: 0.0107 - acc: 0.9968 - val_loss: 0.8651 - val_acc: 0.7074\n",
"Running epoch no. 5\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 32s - loss: 0.0089 - acc: 0.9974 - val_loss: 1.9452 - val_acc: 0.5586\n",
"Running epoch no. 6\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 32s - loss: 0.0086 - acc: 0.9978 - val_loss: 1.5876 - val_acc: 0.6988\n",
"Running epoch no. 7\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 32s - loss: 0.0069 - acc: 0.9982 - val_loss: 1.0250 - val_acc: 0.7863\n",
"Completed 8 fit operations\n"
]
}
],
"source": [
"# Fit the fully connected model\n",
"no_of_epochs=8\n",
"latest_weights_filename = None\n",
"for epoch in range(no_of_epochs):\n",
" print \"Running epoch no. %d\" %epoch\n",
" bn_model.fit(train_features,train_classes,batch_size=batch_size,nb_epoch=1,validation_data=(val_features,val_classes))\n",
" latest_weights_filename = 'bn_0_p06_%d.h5' %epoch\n",
" bn_model.model.save_weights(results_path + latest_weights_filename)\n",
"print \"Completed %s fit operations\" %no_of_epochs"
]
},
{
"cell_type": "code",
"execution_count": 29,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"bn_model.model.load_weights(results_path+'bn_0_p06_0.h5')"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": true
},
"source": [
"## INCLUDE DATA AUGMENTATION\n",
"We want to be able to do data augmentation on the fly and for every run, so we do not want to build a static augmented data set. We can therefore not preprocess the Convolutional block's calcs and feed them to the FC block. We therefore have to combine our bn_model and our conv_layers."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Create a model with integrated convolution and fc layers"
]
},
{
"cell_type": "code",
"execution_count": 24,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"def get_bn_layers(p):\n",
" return [\n",
" MaxPooling2D(input_shape=conv_layers[-1].output_shape[1:]),\n",
" Flatten(),\n",
" Dense(4096,activation='relu'),\n",
" Dropout(p),\n",
" BatchNormalization(),\n",
" Dense(4096,activation='relu'),\n",
" Dropout(p),\n",
" BatchNormalization(),\n",
" Dense(10,activation='softmax')\n",
" ]"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"vgg=Vgg16()\n",
"vgg.ft(10)"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"last_conv_idx=[index for index,layer in enumerate(vgg.model.layers) if type(layer) is Convolution2D][-1]"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"conv_layers = vgg.model.layers[:last_conv_idx+1]"
]
},
{
"cell_type": "code",
"execution_count": 30,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"# Create a final model starting with the Vgg16 convolution block\n",
"final_model = Sequential(conv_layers)\n",
"for layer in final_model.layers: layer.trainable=False\n",
"# Add the bn_layers\n",
"bn_layers = get_bn_layers(0.6)\n",
"for layer in bn_layers: final_model.add(layer)"
]
},
{
"cell_type": "code",
"execution_count": 31,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"bn_model.model.load_weights(results_path+'bn_0_p06_0.h5')"
]
},
{
"cell_type": "code",
"execution_count": 32,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"# Load the weights for the bn-layers from the bn_model\n",
"bn_model.model.load_weights(results_path+'bn_0_p06_0.h5')\n",
"for l1,l2 in zip(final_model.layers[last_conv_idx+1:], bn_model.layers):\n",
" l1.set_weights(l2.get_weights())\n",
"# presumably bn_layers still points to the layers in the final model?"
]
},
{
"cell_type": "code",
"execution_count": 33,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"final_model.compile(optimizer=RMSprop(lr=0.0001),loss='categorical_crossentropy',metrics=['accuracy'])"
]
},
{
"cell_type": "code",
"execution_count": 34,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"final_model.save_weights(results_path+'final0.h5')"
]
},
{
"cell_type": "code",
"execution_count": 51,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"final_model.load_weights(results_path+'final0.h5')"
]
},
{
"cell_type": "code",
"execution_count": 61,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Epoch 1/4\n",
"19977/19977 [==============================] - 588s - loss: 14.5888 - acc: 0.0932 - val_loss: 14.9983 - val_acc: 0.0695\n",
"Epoch 2/4\n",
"19977/19977 [==============================] - 588s - loss: 14.6255 - acc: 0.0926 - val_loss: 14.9983 - val_acc: 0.0695\n",
"Epoch 3/4\n",
"19977/19977 [==============================] - 588s - loss: 14.6303 - acc: 0.0923 - val_loss: 14.9983 - val_acc: 0.0695\n",
"Epoch 4/4\n",
"19977/19977 [==============================] - 588s - loss: 14.6601 - acc: 0.0905 - val_loss: 14.9983 - val_acc: 0.0695\n"
]
},
{
"data": {
"text/plain": [
"<keras.callbacks.History at 0x7fc60bb0ab50>"
]
},
"execution_count": 61,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"final_model.fit_generator(train_batches, samples_per_epoch=train_batches.nb_sample, nb_epoch=4, \n",
" validation_data=val_batches, nb_val_samples=val_batches.nb_sample)"
]
},
{
"cell_type": "code",
"execution_count": 27,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"# Ensure the conv layers has the weights of the VGG model\n",
"source = vgg.model.layers[:last_conv_idx+1]\n",
"target = final_model.layers[:last_conv_idx+1]\n",
"for s_layer,t_layer in zip(source, target):\n",
" t_layer.set_weights(s_layer.get_weights())\n",
" \n",
"# Ensure the conv layers are not trainable\n",
"for layer in final_model.layers[:last_conv_idx+1]:\n",
" layer.trainable = False"
]
},
{
"cell_type": "code",
"execution_count": 28,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"# Ensure the bn layers has the weights of the bn_model\n",
"source = bn_model.layers\n",
"target = final_model.layers[last_conv_idx+1:]\n",
"for s_layer,t_layer in zip(source, target):\n",
" t_layer.set_weights(s_layer.get_weights())"
]
},
{
"cell_type": "code",
"execution_count": 24,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Epoch 1/4\n",
" 2240/19977 [==>...........................] - ETA: 476s - loss: 15.7748 - acc: 0.0045"
]
},
{
"ename": "KeyboardInterrupt",
"evalue": "",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mKeyboardInterrupt\u001b[0m Traceback (most recent call last)",
"\u001b[0;32m<ipython-input-24-16ef47daec02>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m 1\u001b[0m final_model.fit_generator(train_batches, samples_per_epoch=train_batches.nb_sample, nb_epoch=4, \n\u001b[0;32m----> 2\u001b[0;31m validation_data=val_batches, nb_val_samples=val_batches.nb_sample)\n\u001b[0m",
"\u001b[0;32m/home/ubuntu/anaconda2/lib/python2.7/site-packages/keras/models.pyc\u001b[0m in \u001b[0;36mfit_generator\u001b[0;34m(self, generator, samples_per_epoch, nb_epoch, verbose, callbacks, validation_data, nb_val_samples, class_weight, max_q_size, nb_worker, pickle_safe, **kwargs)\u001b[0m\n\u001b[1;32m 872\u001b[0m \u001b[0mmax_q_size\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mmax_q_size\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 873\u001b[0m \u001b[0mnb_worker\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mnb_worker\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 874\u001b[0;31m pickle_safe=pickle_safe)\n\u001b[0m\u001b[1;32m 875\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 876\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mevaluate_generator\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mgenerator\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mval_samples\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmax_q_size\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m10\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnb_worker\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpickle_safe\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mFalse\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m/home/ubuntu/anaconda2/lib/python2.7/site-packages/keras/engine/training.pyc\u001b[0m in \u001b[0;36mfit_generator\u001b[0;34m(self, generator, samples_per_epoch, nb_epoch, verbose, callbacks, validation_data, nb_val_samples, class_weight, max_q_size, nb_worker, pickle_safe)\u001b[0m\n\u001b[1;32m 1441\u001b[0m outs = self.train_on_batch(x, y,\n\u001b[1;32m 1442\u001b[0m \u001b[0msample_weight\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0msample_weight\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1443\u001b[0;31m class_weight=class_weight)\n\u001b[0m\u001b[1;32m 1444\u001b[0m \u001b[0;32mexcept\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1445\u001b[0m \u001b[0m_stop\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mset\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m/home/ubuntu/anaconda2/lib/python2.7/site-packages/keras/engine/training.pyc\u001b[0m in \u001b[0;36mtrain_on_batch\u001b[0;34m(self, x, y, sample_weight, class_weight)\u001b[0m\n\u001b[1;32m 1219\u001b[0m \u001b[0mins\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mx\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0my\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0msample_weights\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1220\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_make_train_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1221\u001b[0;31m \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrain_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mins\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 1222\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutputs\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1223\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0moutputs\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m/home/ubuntu/anaconda2/lib/python2.7/site-packages/keras/backend/theano_backend.pyc\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, inputs)\u001b[0m\n\u001b[1;32m 715\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 716\u001b[0m \u001b[0;32massert\u001b[0m \u001b[0mtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32min\u001b[0m \u001b[0;34m{\u001b[0m\u001b[0mlist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtuple\u001b[0m\u001b[0;34m}\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 717\u001b[0;31m \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfunction\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 718\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 719\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m/home/ubuntu/anaconda2/lib/python2.7/site-packages/theano/compile/function_module.pyc\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m 857\u001b[0m \u001b[0mt0_fn\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtime\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtime\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 858\u001b[0m \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 859\u001b[0;31m \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 860\u001b[0m \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 861\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mhasattr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfn\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'position_of_error'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;31mKeyboardInterrupt\u001b[0m: "
]
}
],
"source": [
"final_model.fit_generator(train_batches, samples_per_epoch=train_batches.nb_sample, nb_epoch=4, \n",
" validation_data=val_batches, nb_val_samples=val_batches.nb_sample)"
]
},
{
"cell_type": "code",
"execution_count": 30,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"vgg.model.optimizer.lr=0.01"
]
},
{
"cell_type": "code",
"execution_count": 31,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Epoch 1/4\n",
"19977/19977 [==============================] - 560s - loss: 14.4394 - acc: 0.1020 - val_loss: 14.4845 - val_acc: 0.1013\n",
"Epoch 2/4\n",
"19977/19977 [==============================] - 562s - loss: 14.4254 - acc: 0.1050 - val_loss: 14.4845 - val_acc: 0.1013\n",
"Epoch 3/4\n",
"19977/19977 [==============================] - 563s - loss: 14.4254 - acc: 0.1050 - val_loss: 14.4845 - val_acc: 0.1013\n",
"Epoch 4/4\n",
"19977/19977 [==============================] - 562s - loss: 14.4254 - acc: 0.1050 - val_loss: 14.4845 - val_acc: 0.1013\n"
]
},
{
"data": {
"text/plain": [
"<keras.callbacks.History at 0x7f9410e46990>"
]
},
"execution_count": 31,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"vgg.model.fit_generator(train_batches, samples_per_epoch=train_batches.nb_sample, nb_epoch=4, \n",
" validation_data=val_batches, nb_val_samples=val_batches.nb_sample)"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"for i in range(len(vgg.model.layers[last_conv_idx+1+2:])):\n",
" vgg.model.pop()"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"bn_layers = get_bn_layers(0.6)"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"for layer in bn_layers:\n",
" vgg.model.add(layer)"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"for layer in vgg.model.layers[:last_conv_idx+1]:\n",
" layer.trainable = False"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"bn_model.model.load_weights(results_path+'bn_0_p06_0.h5')"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"for l1,l2 in zip(vgg.model.layers[last_conv_idx+1:], bn_model.layers):\n",
" l1.set_weights(l2.get_weights())"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"vgg.compile()"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Epoch 1/1\n",
"19977/19977 [==============================] - 594s - loss: 2.4881 - acc: 0.3449 - val_loss: 2.8662 - val_acc: 0.0830\n"
]
}
],
"source": [
"vgg.fit(train_batches,val_batches)"
]
},
{
"cell_type": "code",
"execution_count": 24,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"vgg.model.compile(optimizer=RMSprop(lr=0.0001),loss='categorical_crossentropy',metrics=['accuracy'])"
]
},
{
"cell_type": "code",
"execution_count": 25,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Epoch 1/1\n",
" 2624/19977 [==>...........................] - ETA: 464s - loss: 4.5664 - acc: 0.0114"
]
},
{
"ename": "KeyboardInterrupt",
"evalue": "",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mKeyboardInterrupt\u001b[0m Traceback (most recent call last)",
"\u001b[0;32m<ipython-input-25-74a0b444a8ce>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mvgg\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtrain_batches\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mval_batches\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
"\u001b[0;32m/home/ubuntu/courses/deeplearning1/nbs/vgg16.pyc\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, batches, val_batches, nb_epoch)\u001b[0m\n\u001b[1;32m 115\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatches\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mval_batches\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnb_epoch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 116\u001b[0m self.model.fit_generator(batches, samples_per_epoch=batches.nb_sample, nb_epoch=nb_epoch,\n\u001b[0;32m--> 117\u001b[0;31m validation_data=val_batches, nb_val_samples=val_batches.nb_sample)\n\u001b[0m\u001b[1;32m 118\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 119\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m/home/ubuntu/anaconda2/lib/python2.7/site-packages/keras/models.pyc\u001b[0m in \u001b[0;36mfit_generator\u001b[0;34m(self, generator, samples_per_epoch, nb_epoch, verbose, callbacks, validation_data, nb_val_samples, class_weight, max_q_size, nb_worker, pickle_safe, **kwargs)\u001b[0m\n\u001b[1;32m 872\u001b[0m \u001b[0mmax_q_size\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mmax_q_size\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 873\u001b[0m \u001b[0mnb_worker\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mnb_worker\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 874\u001b[0;31m pickle_safe=pickle_safe)\n\u001b[0m\u001b[1;32m 875\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 876\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mevaluate_generator\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mgenerator\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mval_samples\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmax_q_size\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m10\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnb_worker\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpickle_safe\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mFalse\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m/home/ubuntu/anaconda2/lib/python2.7/site-packages/keras/engine/training.pyc\u001b[0m in \u001b[0;36mfit_generator\u001b[0;34m(self, generator, samples_per_epoch, nb_epoch, verbose, callbacks, validation_data, nb_val_samples, class_weight, max_q_size, nb_worker, pickle_safe)\u001b[0m\n\u001b[1;32m 1441\u001b[0m outs = self.train_on_batch(x, y,\n\u001b[1;32m 1442\u001b[0m \u001b[0msample_weight\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0msample_weight\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1443\u001b[0;31m class_weight=class_weight)\n\u001b[0m\u001b[1;32m 1444\u001b[0m \u001b[0;32mexcept\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1445\u001b[0m \u001b[0m_stop\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mset\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m/home/ubuntu/anaconda2/lib/python2.7/site-packages/keras/engine/training.pyc\u001b[0m in \u001b[0;36mtrain_on_batch\u001b[0;34m(self, x, y, sample_weight, class_weight)\u001b[0m\n\u001b[1;32m 1219\u001b[0m \u001b[0mins\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mx\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0my\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0msample_weights\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1220\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_make_train_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1221\u001b[0;31m \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrain_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mins\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 1222\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutputs\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1223\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0moutputs\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m/home/ubuntu/anaconda2/lib/python2.7/site-packages/keras/backend/theano_backend.pyc\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, inputs)\u001b[0m\n\u001b[1;32m 715\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 716\u001b[0m \u001b[0;32massert\u001b[0m \u001b[0mtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32min\u001b[0m \u001b[0;34m{\u001b[0m\u001b[0mlist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtuple\u001b[0m\u001b[0;34m}\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 717\u001b[0;31m \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfunction\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 718\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 719\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m/home/ubuntu/anaconda2/lib/python2.7/site-packages/theano/compile/function_module.pyc\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m 857\u001b[0m \u001b[0mt0_fn\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtime\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtime\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 858\u001b[0m \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 859\u001b[0;31m \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 860\u001b[0m \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 861\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mhasattr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfn\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'position_of_error'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;31mKeyboardInterrupt\u001b[0m: "
]
}
],
"source": [
"vgg.fit(train_batches,val_batches)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"no_of_epochs=5\n",
"latest_weights_filename = None\n",
"for epoch in range(no_of_epochs):\n",
" print \"Running epoch no. %d\" %epoch\n",
" bn_model.fit(train_features,train_classes,batch_size=batch_size,nb_epoch=1,validation_data=(val_features,val_classes))\n",
" latest_weights_filename = 'bn_0_p06_%d.h5' %epoch\n",
" bn_model.model.save_weights(results_path + latest_weights_filename)\n",
"print \"Completed %s fit operations\" %no_of_epochs"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## EVALUATE DATA AND MODELS"
]
},
{
"cell_type": "code",
"execution_count": 35,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"i=0"
]
},
{
"cell_type": "code",
"execution_count": 43,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"a_batch = train_batches.next()"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"2"
]
},
"execution_count": 20,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"len(a_batch)"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"64"
]
},
"execution_count": 21,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"len(a_batch[0])"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"64"
]
},
"execution_count": 22,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"len(a_batch[1])"
]
},
{
"cell_type": "code",
"execution_count": 42,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAWoAAAFjCAYAAAAU10ErAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ncZeiW0wAAAAASUVORK5CYII=\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f0cb19b77d0>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"plot(a_batch[0][0])\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 134,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"bn_model.model.load_weights(results_path+'bn_0_p06_0.h5')"
]
},
{
"cell_type": "code",
"execution_count": 135,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"final_model.load_weights(results_path+'final0.h5')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"# Set final_model weights equal to bn_model weights (fc layers only)\n",
"source = bn_model.layers\n",
"target = final_model.layers[last_conv_idx+1:]\n",
"for l_s,l_t in zip(source,target):\n",
" l_t.set_weights(l_s.get_weights())"
]
},
{
"cell_type": "code",
"execution_count": 37,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"# Set fc_model weights equal to vgg model weights (fc layers only)\n",
"source = vgg.model.layers[last_conv_idx+1:]\n",
"target = fc_model_lst.layers\n",
"for l_s,l_t in zip(source,target):\n",
" l_t.set_weights(l_s.get_weights())"
]
},
{
"cell_type": "code",
"execution_count": 44,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"64/64 [==============================] - 0s \n"
]
},
{
"data": {
"text/plain": [
"[3.2040026187896729, 0.046875]"
]
},
"execution_count": 44,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"fc_model_lst.evaluate(train_features[i:i+64],train_classes[i:i+64])"
]
},
{
"cell_type": "code",
"execution_count": 45,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"64/64 [==============================] - 1s \n"
]
},
{
"data": {
"text/plain": [
"[3.2040026187896729, 0.046875]"
]
},
"execution_count": 45,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"vgg.model.evaluate(a_batch[0],a_batch[1])"
]
},
{
"cell_type": "code",
"execution_count": 46,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"64/64 [==============================] - 1s \n"
]
},
{
"data": {
"text/plain": [
"[0.020014840876683593, 0.984375]"
]
},
"execution_count": 46,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"final_model.evaluate(a_batch[0],a_batch[1])"
]
},
{
"cell_type": "code",
"execution_count": 47,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"64/64 [==============================] - 0s \n"
]
},
{
"data": {
"text/plain": [
"[0.020014840876683593, 0.984375]"
]
},
"execution_count": 47,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"bn_model.evaluate(train_features[i:i+64],train_classes[i:i+64])"
]
},
{
"cell_type": "code",
"execution_count": 49,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"True"
]
},
"execution_count": 49,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"vgg.model.layers[:last_conv_idx+1]==final_model.layers[:last_conv_idx+1]"
]
},
{
"cell_type": "code",
"execution_count": 52,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"bn_model.optimizer.lr = .0001"
]
},
{
"cell_type": "code",
"execution_count": 106,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Running epoch no. 0\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 31s - loss: 0.0400 - acc: 0.9877 - val_loss: 0.7098 - val_acc: 0.7752\n",
"Running epoch no. 1\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 32s - loss: 0.0218 - acc: 0.9936 - val_loss: 0.7958 - val_acc: 0.7589\n",
"Running epoch no. 2\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 32s - loss: 0.0176 - acc: 0.9946 - val_loss: 1.1043 - val_acc: 0.6812\n",
"Running epoch no. 3\n",
"Train on 19977 samples, validate on 2447 samples\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 32s - loss: 0.0109 - acc: 0.9966 - val_loss: 1.1416 - val_acc: 0.6669\n",
"Completed 4 fit operations\n"
]
}
],
"source": [
"no_of_epochs=4\n",
"latest_weights_filename = None\n",
"for epoch in range(no_of_epochs):\n",
" print \"Running epoch no. %d\" %epoch\n",
" bn_model.fit(train_features,train_classes,batch_size=batch_size,nb_epoch=1,validation_data=(val_features,val_classes))\n",
" latest_weights_filename = 'bn_1_p06_%d.h5' %epoch\n",
" bn_model.model.save_weights(results_path + latest_weights_filename)\n",
"print \"Completed %s fit operations\" %no_of_epochs"
]
},
{
"cell_type": "code",
"execution_count": 55,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"final_model.optimizer.lr = .0001"
]
},
{
"cell_type": "code",
"execution_count": 56,
"metadata": {
"collapsed": false,
"scrolled": true
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Running epoch no. 0\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 582s - loss: 3.1267 - acc: 0.1120 - val_loss: 1.7230 - val_acc: 0.5333\n",
"Running epoch no. 1\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 583s - loss: 2.7573 - acc: 0.1024 - val_loss: 2.1196 - val_acc: 0.3347\n",
"Running epoch no. 2\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 584s - loss: 2.6088 - acc: 0.1061 - val_loss: 2.2088 - val_acc: 0.2554\n",
"Running epoch no. 3\n",
"Epoch 1/1\n",
"19977/19977 [==============================] - 584s - loss: 2.5073 - acc: 0.1023 - val_loss: 2.2339 - val_acc: 0.2930\n",
"Completed 4 fit operations\n"
]
}
],
"source": [
"no_of_epochs=4\n",
"latest_weights_filename = None\n",
"for epoch in range(no_of_epochs):\n",
" print \"Running epoch no. %d\" %epoch\n",
" final_model.fit_generator(train_batches, samples_per_epoch=train_batches.nb_sample, nb_epoch=1, \n",
" validation_data=val_batches, nb_val_samples=val_batches.nb_sample)\n",
" latest_weights_filename = 'f_1_p06_%d.h5' %epoch\n",
" bn_model.model.save_weights(results_path + latest_weights_filename)\n",
"print \"Completed %s fit operations\" %no_of_epochs"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Check that the final_model convolution weights did not change from the original Vgg16 model weights"
]
},
{
"cell_type": "code",
"execution_count": 132,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"All convolution layers are equal: True\n"
]
}
],
"source": [
"test = True\n",
"v_layers = vgg.model.layers[:last_conv_idx+1]\n",
"f_layers = final_model.layers[:last_conv_idx+1]\n",
"for vl,fl in zip(v_layers,f_layers):\n",
" for i in range(len(vl.get_weights())):\n",
" test = test & (vl.get_weights()[i]==fl.get_weights()[i]).all()\n",
"print 'All convolution layers are equal: %s' %test"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Check wether weights of the fc layers in the bn_model and final_model are the same before training"
]
},
{
"cell_type": "code",
"execution_count": 136,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"All fc layers are equal: True\n"
]
}
],
"source": [
"test = True\n",
"b_layers = bn_model.layers\n",
"f_layers = final_model.layers[last_conv_idx+1:]\n",
"for bl,fl in zip(b_layers,f_layers):\n",
" for i in range(len(bl.get_weights())):\n",
" test=test&(bl.get_weights()[i]==fl.get_weights()[i]).all()\n",
"print 'All fc layers are equal: %s' %test"
]
}
],
"metadata": {
"anaconda-cloud": {},
"kernelspec": {
"display_name": "Python [conda root]",
"language": "python",
"name": "conda-root-py"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 2
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython2",
"version": "2.7.12"
}
},
"nbformat": 4,
"nbformat_minor": 1
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment