dcgan_autoencoder_notebook.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using gpu device 0: Graphics Device (CNMeM is disabled)\n"
     ]
    }
   ],
   "source": [
    "import sys\n",
    "sys.path.append('..')\n",
    "\n",
    "import os\n",
    "import json\n",
    "from time import time\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "\n",
    "import theano\n",
    "import theano.tensor as T\n",
    "from theano.sandbox.cuda.dnn import dnn_conv\n",
    "\n",
    "from PIL import Image"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "N.B. The code from the following imports is lifted from the original [dcgan project](https://github.com/Newmu/dcgan_code)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from lib import activations\n",
    "from lib import updates\n",
    "from lib import inits\n",
    "from lib.rng import py_rng, np_rng\n",
    "from lib.ops import batchnorm, conv_cond_concat, deconv, dropout, l2normalize\n",
    "from lib.metrics import nnc_score, nnd_score\n",
    "from lib.theano_utils import floatX, sharedX\n",
    "from lib.data_utils import OneHot, shuffle, iter_data, center_crop, patch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from fuel.datasets.hdf5 import H5PYDataset\n",
    "from fuel.schemes import ShuffledScheme, SequentialScheme\n",
    "from fuel.streams import DataStream"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Stuff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number of samples in dataset : 9216\n"
     ]
    }
   ],
   "source": [
    "import h5py\n",
    "try:\n",
    "    hf[\"target\"].shape\n",
    "except:\n",
    "    hf = h5py.File('faces.hdf5','r+')\n",
    "num_samples = hf[\"input\"].shape[0]\n",
    "\n",
    "print \"number of samples in dataset : %i\" %num_samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "split_dict = {\n",
    "     'train': {'input': (2000, num_samples), 'target': (2000, num_samples)},\n",
    "     'test': {'input': (0, 1000), 'target': (0, 1000)},\n",
    "     'val': {'input': (1000, 2000), 'target': (1000, 2000)}\n",
    "}\n",
    "hf.attrs['split'] = H5PYDataset.create_split_array(split_dict)\n",
    "train_set = H5PYDataset('faces.hdf5', which_sets=('train',))\n",
    "test_set = H5PYDataset('faces.hdf5', which_sets=('test',))\n",
    "val_set = H5PYDataset('faces.hdf5', which_sets=('val',))\n",
    "\n",
    "batch_size = 128\n",
    "\n",
    "#TODO : use shuffledscheme instead?  Seems slower, might have screwed up the chunksize in the HDF5 files?\n",
    "\n",
    "tr_scheme = SequentialScheme(examples=train_set.num_examples, batch_size=batch_size)\n",
    "tr_stream = DataStream(train_set, iteration_scheme=tr_scheme)\n",
    "\n",
    "val_scheme = SequentialScheme(examples=val_set.num_examples, batch_size=batch_size)\n",
    "val_stream = DataStream(val_set, iteration_scheme=val_scheme)\n",
    "\n",
    "test_scheme = SequentialScheme(examples=test_set.num_examples, batch_size=batch_size)\n",
    "test_stream = DataStream(test_set, iteration_scheme=test_scheme)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Check data looks sensible"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "EXAMPLE TARGET IMAGE:\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAIAAAACgCAIAAABL8POqAAB9tklEQVR4nOz9aZRl2XUeiO29z3Cn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\nKvP1pRXfmr6UUjLz/wJKendXs2iAxQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<PIL.Image.Image image mode=RGB size=128x160 at 0x7FEEBA533750>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "for x_train, x_target in tr_stream.get_epoch_iterator():\n",
    "    break\n",
    "print \"EXAMPLE TARGET IMAGE:\"\n",
    "\n",
    "Image.fromarray(x_target[3].astype(np.uint8))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "EXAMPLE INPUT IMAGE:\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAACAAAAAoCAIAAAAQS2/OAAAK+klEQVR4nIVWWYxl11Xd+5xzpzfc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\nmM8oV74AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<PIL.Image.Image image mode=RGB size=32x40 at 0x7FEEB84DB110>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print \"EXAMPLE INPUT IMAGE:\"\n",
    "Image.fromarray(x_train[3].astype(np.uint8))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Setup Neural Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def target_transform(X):\n",
    "    return floatX(X).transpose(0, 3, 1, 2)/127.5 - 1.\n",
    "\n",
    "def input_transform(X):\n",
    "    return target_transform(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "l2 = 1e-5         # l2 weight decay\n",
    "nvis = 196        # # of samples to visualize during training\n",
    "b1 = 0.5          # momentum term of adam\n",
    "nc = 3            # # of channels in image\n",
    "nbatch = 128      # # of examples in batch\n",
    "npx = 64          # # of pixels width/height of images\n",
    "\n",
    "nx = npx*npx*nc   # # of dimensions in X\n",
    "niter = 1000        # # of iter at starting learning rate\n",
    "niter_decay = 30   # # of iter to linearly decay learning rate to zero\n",
    "lr = 0.0002       # initial learning rate for adam\n",
    "ntrain = 25000   # # of examples to train on\n",
    "\n",
    "relu = activations.Rectify()\n",
    "sigmoid = activations.Sigmoid()\n",
    "lrelu = activations.LeakyRectify()\n",
    "tanh = activations.Tanh()\n",
    "bce = T.nnet.binary_crossentropy\n",
    "\n",
    "def mse(x,y):\n",
    "    return T.sum(T.pow(x-y,2), axis = 1)\n",
    "\n",
    "gifn = inits.Normal(scale=0.02)\n",
    "difn = inits.Normal(scale=0.02)\n",
    "sigma_ifn = inits.Normal(loc = -100., scale=0.02)\n",
    "gain_ifn = inits.Normal(loc=1., scale=0.02)\n",
    "bias_ifn = inits.Constant(c=0.)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "The following methods are to help adjust the sizes of the convolutional layers in the generator and discriminator, which is very fiddly to do otherwise.  The (overloaded) method make_conv_set can be used to create both the conv \n",
    "and deconv sets of layers.  Note that the 'size' of the images is the size of the shortest side (32 in the input set, 128 in the target set).  Only use powers of 2 here."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def make_conv_layer(X, input_size, output_size, input_filters, \n",
    "                    output_filters, name, index,\n",
    "                    weights = None, filter_sz = 5):\n",
    "    \n",
    "    is_deconv = output_size >= input_size\n",
    "\n",
    "    w_size = (input_filters, output_filters, filter_sz, filter_sz) \\\n",
    "            if is_deconv else (output_filters, input_filters, filter_sz, filter_sz)\n",
    "    \n",
    "    if weights is None:\n",
    "        w = gifn(w_size, '%sw%i' %(name, index))\n",
    "        g = gain_ifn((output_filters), '%sg%i' %(name, index))\n",
    "        b = bias_ifn((output_filters), '%sb%i' %(name, index))\n",
    "    else:\n",
    "        w,g,b = weights\n",
    "    \n",
    "    conv_method = deconv if is_deconv else dnn_conv\n",
    "    activation = relu if is_deconv else lrelu\n",
    "    \n",
    "    sub = output_size / input_size if is_deconv else input_size / output_size\n",
    "    \n",
    "    if filter_sz == 3:\n",
    "        bm = 1\n",
    "    else:\n",
    "        bm = 2\n",
    "    \n",
    "    layer = activation(batchnorm(conv_method(X, w, subsample=(sub, sub), border_mode=(bm, bm)), g=g, b=b))\n",
    "    \n",
    "    return layer, [w,g,b]\n",
    "\n",
    "def make_conv_set(input, layer_sizes, num_filters, name, weights = None, filter_szs = None):\n",
    "    assert(len(layer_sizes) == len(num_filters))\n",
    "    \n",
    "    vars_ = []\n",
    "    layers_ = []\n",
    "    current_layer = input\n",
    "    \n",
    "    for i in range(len(layer_sizes) - 1):\n",
    "        input_size = layer_sizes[i]\n",
    "        output_size = layer_sizes[i + 1]\n",
    "        input_filters = num_filters[i]\n",
    "        output_filters = num_filters[i + 1]\n",
    "        \n",
    "        if weights is not None:\n",
    "            this_wts = weights[i * 3 : i * 3 + 3]\n",
    "        else:\n",
    "            this_wts = None\n",
    "            \n",
    "        if filter_szs != None:\n",
    "            filter_sz = filter_szs[i]\n",
    "        else:\n",
    "            filter_sz = 5\n",
    "        \n",
    "        layer, new_vars = make_conv_layer(current_layer, input_size, output_size, \n",
    "                                          input_filters, output_filters, name, i, \n",
    "                                          weights = this_wts, filter_sz = filter_sz)\n",
    "        \n",
    "        vars_ += new_vars\n",
    "        layers_ += [layer]\n",
    "        current_layer = layer\n",
    "        \n",
    "    return current_layer, vars_, layers_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# inputs\n",
    "X = T.tensor4()\n",
    "\n",
    "## encode layer\n",
    "e_layer_sizes = [32, 16, 8]\n",
    "e_filter_sizes = [3, 256, 1024]\n",
    "\n",
    "eX, e_params, e_layers = make_conv_set(X, e_layer_sizes, e_filter_sizes, \"e\")\n",
    "\n",
    "## generative layer\n",
    "g_layer_sizes = [8, 16, 32, 64, 128]\n",
    "g_num_filters = [1024, 512, 256, 256, 128]\n",
    "\n",
    "g_out, g_params, g_layers = make_conv_set(eX, g_layer_sizes, g_num_filters, \"g\")\n",
    "gwx = gifn((128, nc, 5, 5), 'gwx')\n",
    "g_params += [gwx]\n",
    "gX = tanh(deconv(g_out, gwx, subsample=(1, 1), border_mode=(2, 2)))\n",
    "\n",
    "## discrim layer(s)\n",
    "\n",
    "df1 = 128\n",
    "d_layer_sizes = [128, 64, 32, 16, 8]\n",
    "d_filter_sizes = [3, df1, 2 * df1, 4 * df1, 8 * df1]\n",
    "\n",
    "dwy = difn((df1 * 8 * 10 * 8, 1), 'dwy')\n",
    "\n",
    "def discrim(input, name, weights=None):\n",
    "    d_out, disc_params, d_layers = make_conv_set(input, d_layer_sizes, d_filter_sizes, name, weights = weights)\n",
    "    d_flat = T.flatten(d_out, 2)\n",
    "    \n",
    "    disc_params += [dwy]\n",
    "    y = sigmoid(T.dot(d_flat, dwy))\n",
    "    \n",
    "    return y, disc_params, d_layers\n",
    "\n",
    "# target outputs\n",
    "target = T.tensor4()\n",
    "\n",
    "p_real, d_params, d_layers = discrim(target, \"d\")\n",
    "#we need to make sure the p_gen params are the same as the p_real params\n",
    "p_gen , d_params2, d_layers = discrim(gX, \"d\", weights=d_params)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(128, 1)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# test everything working so far (errors are most likely size mismatches)\n",
    "f = theano.function([X], p_gen)\n",
    "f(input_transform(x_train)).shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next we set up the various cost functions we need"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from theano.tensor.signal.downsample import max_pool_2d\n",
    "\n",
    "## GAN costs\n",
    "d_cost_real = bce(p_real, T.ones(p_real.shape)).mean()\n",
    "d_cost_gen = bce(p_gen, T.zeros(p_gen.shape)).mean()\n",
    "g_cost_d = bce(p_gen, T.ones(p_gen.shape)).mean()\n",
    "\n",
    "## MSE encoding cost is done on an (averaged) downscaling of the image\n",
    "target_pool = max_pool_2d(target, (4,4), mode=\"average_exc_pad\",ignore_border=True)\n",
    "target_flat = T.flatten(target_pool, 2)\n",
    "gX_pool = max_pool_2d(gX, (4,4), mode=\"average_exc_pad\",ignore_border=True)\n",
    "gX_flat = T.flatten(gX_pool,2)\n",
    "enc_cost = mse(gX_flat, target_flat).mean() \n",
    "\n",
    "## generator cost is a linear combination of the discrim cost plus the MSE enocding cost\n",
    "d_cost = d_cost_real + d_cost_gen\n",
    "g_cost = g_cost_d + enc_cost / 100   ## if the enc_cost is weighted too highly it will take a long time to train\n",
    "\n",
    "## N.B. e_cost and e_updates will only try and minimise MSE loss on the autoencoder (for debugging)\n",
    "e_cost = enc_cost\n",
    "\n",
    "cost = [g_cost_d, d_cost_real, enc_cost]\n",
    "\n",
    "elrt = sharedX(0.002)\n",
    "lrt = sharedX(lr)\n",
    "d_updater = updates.Adam(lr=lrt, b1=b1, regularizer=updates.Regularizer(l2=l2))\n",
    "g_updater = updates.Adam(lr=lrt, b1=b1, regularizer=updates.Regularizer(l2=l2))\n",
    "e_updater = updates.Adam(lr=elrt, b1=b1, regularizer=updates.Regularizer(l2=l2))\n",
    "\n",
    "d_updates = d_updater(d_params, d_cost)\n",
    "g_updates = g_updater(e_params + g_params, g_cost)\n",
    "e_updates = e_updater(e_params, e_cost)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "COMPILING\n",
      "30.91 seconds to compile theano functions\n"
     ]
    }
   ],
   "source": [
    "print 'COMPILING'\n",
    "t = time()\n",
    "_train_g = theano.function([X, target], cost, updates=g_updates)\n",
    "_train_d = theano.function([X, target], cost, updates=d_updates)\n",
    "_train_e = theano.function([X, target], cost, updates=e_updates)\n",
    "_get_cost = theano.function([X, target], cost)\n",
    "print '%.2f seconds to compile theano functions'%(time()-t)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training code\n",
    "\n",
    "Code for generating the images every 100 batches or so."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "img_dir = \"gen_images/\"\n",
    "\n",
    "if not os.path.exists(img_dir):\n",
    "    os.makedirs(img_dir)\n",
    "\n",
    "ae_encode = theano.function([X, target], [gX, target])\n",
    "\n",
    "def inverse(X):\n",
    "    X_pred = (X.transpose(0, 2, 3, 1) + 1) * 127.5\n",
    "    X_pred = np.rint(X_pred).astype(int)\n",
    "    X_pred = np.clip(X_pred, a_min = 0, a_max = 255)\n",
    "    return X_pred.astype('uint8')\n",
    "\n",
    "\n",
    "def save_sample_pictures():\n",
    "    for te_train, te_target in test_stream.get_epoch_iterator():\n",
    "        break\n",
    "    te_out, te_ta = ae_encode(input_transform(te_train), target_transform(te_target))\n",
    "    te_reshape = inverse(te_out)\n",
    "    te_target_reshape = inverse(te_ta)\n",
    "\n",
    "    new_size = (128 * 6, 160 * 12)\n",
    "    new_im = Image.new('RGB', new_size)\n",
    "    r = np.random.choice(128, 24, replace=False).reshape(2,12)\n",
    "    for i in range(2):\n",
    "        for j in range(12):\n",
    "            index =  r[i][j]\n",
    "            \n",
    "            target_im = Image.fromarray(te_target_reshape[index])\n",
    "            train_im = Image.fromarray(te_train[index].repeat(axis=0,repeats=4).repeat(axis=1,repeats=4).astype(np.uint8))\n",
    "            im = Image.fromarray(te_reshape[index])\n",
    "            \n",
    "            new_im.paste(target_im, (128 * i * 3, 160 * j))\n",
    "            new_im.paste(train_im, (128 * (i * 3 + 1), 160 * j))\n",
    "            new_im.paste(im, (128 * (i * 3 + 2), 160 * j))\n",
    "    img_loc = \"gen_images/%i.png\" %int(time())     \n",
    "    print \"saving images to %s\" %img_loc\n",
    "    new_im.save(img_loc)\n",
    "    \n",
    "save_sample_pictures()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def mn(l):\n",
    "    if sum(l) == 0:\n",
    "        return 0\n",
    "    return sum(l) / len(l)\n",
    "\n",
    "## TODO : nicer way of coding these means?\n",
    "\n",
    "def get_test_errors():\n",
    "    print \"getting test error\"\n",
    "    g_costs = []\n",
    "    d_costs = []\n",
    "    e_costs = []\n",
    "    k_costs = []\n",
    "    for i in range(20):\n",
    "        try:\n",
    "            x_train, x_target  = te_iterator.next()\n",
    "        except:\n",
    "            te_iterator = val_stream.get_epoch_iterator()\n",
    "            x_train, x_target  = te_iterator.next()\n",
    "        x = input_transform(x_train)\n",
    "        t = target_transform(x_target)\n",
    "        cost = _get_cost(x,t)\n",
    "        g_cost, d_cost, enc_cost = cost\n",
    "        g_costs.append(g_cost)\n",
    "        d_costs.append(d_cost)\n",
    "        e_costs.append(enc_cost)\n",
    "        \n",
    "    s= \" ,\".join([\"test errors :\", str(mn(g_costs)), str(mn(d_costs)), str(mn(e_costs))])\n",
    "    return s\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train Model\n",
    "\n",
    "Finally, we come to the actual training of the model.  This code can be keyboard interrupted, and the weights will be stored in memory, allowing us to stop, adjust and restart the training (this is how I got the model to train).  For advice on training see the blog post at (#TODO)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "iterator = tr_stream.get_epoch_iterator()\n",
    "\n",
    "# you may wish to reset the learning rate to something of your choosing if you feel it is too high/low\n",
    "lrt = sharedX(lr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/72 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "STARTING\n",
      "getting test error\n",
      "test errors : ,48.5995481491 ,3.94193813801 ,645.511212158\n",
      "training errors : ,6.17664718628 ,6.20431184769 ,654.965026855\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                               "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 0 of 200 took 308.92 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 39%|███▉      | 28/72 [01:43<02:43,  0.27it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "getting test error\n",
      "test errors : ,8.7343829155 ,4.93261101246 ,168.448952484\n",
      "training errors : ,4.3123496329 ,1.67833978529 ,178.185795751\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                               "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 1 of 200 took 301.45 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 78%|███████▊  | 56/72 [03:26<00:59,  0.27it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "getting test error\n",
      "test errors : ,4.42765614986 ,2.67707374096 ,119.657383728\n",
      "training errors : ,2.4958941113 ,1.4228379116 ,147.280639648\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                               "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 2 of 200 took 300.45 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                               "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 3 of 200 took 265.26 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 17%|█▋        | 12/72 [00:44<03:41,  0.27it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "getting test error\n",
      "test errors : ,5.20457861423 ,2.05324191451 ,116.56504097\n",
      "training errors : ,2.26672356862 ,1.0022797596 ,117.856253991\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 36%|███▌      | 26/72 [02:11<02:50,  0.27it/s]"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-33-65f523c7e6dd>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m     32\u001b[0m             \u001b[0mcost\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0m_train_g\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mt\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     33\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 34\u001b[1;33m             \u001b[0mcost\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0m_train_d\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mt\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     35\u001b[0m         \u001b[1;31m#cost = _train_e(x,t)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     36\u001b[0m         \u001b[0mg_cost\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0md_cost\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0menc_cost\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcost\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m/home/mike/envs/lasagne/lib/python2.7/site-packages/theano/compile/function_module.pyc\u001b[0m in \u001b[0;36m__call__\u001b[1;34m(self, *args, **kwargs)\u001b[0m\n\u001b[0;32m    857\u001b[0m         \u001b[0mt0_fn\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtime\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtime\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    858\u001b[0m         \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 859\u001b[1;33m             \u001b[0moutputs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfn\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    860\u001b[0m         \u001b[1;32mexcept\u001b[0m \u001b[0mException\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    861\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0mhasattr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfn\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'position_of_error'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "from time import time\n",
    "\n",
    "n_updates = 0\n",
    "t = time()\n",
    "\n",
    "n_epochs = 200\n",
    "\n",
    "print \"STARTING\"\n",
    "\n",
    "\n",
    "\n",
    "for epoch in range(n_epochs):\n",
    "    \n",
    "    \n",
    "    tm = time()\n",
    "\n",
    "    g_costs = []\n",
    "    d_costs = []\n",
    "    e_costs = []\n",
    "    \n",
    "    ## TODO : produces pretty ugly output, redo this?\n",
    "    for i in tqdm(range(num_samples/128)):\n",
    "        \n",
    "        try:\n",
    "            x_train, x_target  = iterator.next()\n",
    "        except:\n",
    "            iterator = tr_stream.get_epoch_iterator()\n",
    "            x_train, x_target  = iterator.next()\n",
    "        x = input_transform(x_train)\n",
    "        t = target_transform(x_target)\n",
    "\n",
    "        ## optional - change the criteria for how often we train the generator or discriminator\n",
    "        if n_updates % 2 == 1:\n",
    "            cost = _train_g(x,t) \n",
    "        else:\n",
    "            cost = _train_d(x,t)\n",
    "            \n",
    "        # optional - only train the generator on MSE cost\n",
    "        #cost = _train_e(x,t)\n",
    "        g_cost, d_cost, enc_cost = cost\n",
    "        g_costs.append(g_cost)\n",
    "        d_costs.append(d_cost)\n",
    "        e_costs.append(enc_cost)\n",
    "\n",
    "        if n_updates % 100 == 0:\n",
    "            s= \" ,\".join([\"training errors :\", str(mn(g_costs)), str(mn(d_costs)), str(mn(e_costs))])\n",
    "            g_costs = []\n",
    "            d_costs = []\n",
    "            e_costs = []\n",
    "            print get_test_errors()\n",
    "            print s\n",
    "            sys.stdout.flush()\n",
    "            save_sample_pictures()\n",
    "        n_updates += 1  \n",
    "\n",
    "    print \"epoch %i of %i took %.2f seconds\" %(epoch, n_epochs, time() - tm)\n",
    "    \n",
    "    ## optional - reduce the learning rate as you go\n",
    "    #lrt.set_value(floatX(lrt.get_value() * 0.95))\n",
    "    #print lrt.get_value()\n",
    "    \n",
    "    \n",
    "    sys.stdout.flush()\n",
    "    \n",
    "    \n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Save weights if wanted\n",
    "You can reuse them by using the weights in the make_conv_set method #TODO - actually try this!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import pickle\n",
    "\n",
    "all_params = [e_params, g_params, d_params]\n",
    "\n",
    "pickle.dump(all_params, open(\"faces_dcgan.pkl\", 'w'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}