ex_5.6_2_auto_encoders_deconv.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting data/MNIST/train-images-idx3-ubyte.gz\n",
      "Extracting data/MNIST/train-labels-idx1-ubyte.gz\n",
      "Extracting data/MNIST/t10k-images-idx3-ubyte.gz\n",
      "Extracting data/MNIST/t10k-labels-idx1-ubyte.gz\n"
     ]
    },
    {
     "ename": "InvalidArgumentError",
     "evalue": "Conv2DCustomBackpropInput: input and out_backprop must have the same batch sizeinput batch: 64outbackprop batch: 10000 batch_dim: 0\n\t [[Node: conv2d_transpose_2 = Conv2DBackpropInput[T=DT_FLOAT, data_format=\"NHWC\", dilations=[1, 1, 1, 1], padding=\"SAME\", strides=[1, 2, 2, 1], use_cudnn_on_gpu=true, _device=\"/job:localhost/replica:0/task:0/device:CPU:0\"](stack_4, Variable_22/read, Sigmoid_4)]]\n\nCaused by op 'conv2d_transpose_2', defined at:\n  File \"/usr/lib/python3.6/runpy.py\", line 193, in _run_module_as_main\n    \"__main__\", mod_spec)\n  File \"/usr/lib/python3.6/runpy.py\", line 85, in _run_code\n    exec(code, run_globals)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/ipykernel_launcher.py\", line 16, in <module>\n    app.launch_new_instance()\n  File \"/home/bekin/.local/lib/python3.6/site-packages/traitlets/config/application.py\", line 658, in launch_instance\n    app.start()\n  File \"/home/bekin/.local/lib/python3.6/site-packages/ipykernel/kernelapp.py\", line 478, in start\n    self.io_loop.start()\n  File \"/home/bekin/.local/lib/python3.6/site-packages/zmq/eventloop/ioloop.py\", line 177, in start\n    super(ZMQIOLoop, self).start()\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tornado/ioloop.py\", line 888, in start\n    handler_func(fd_obj, events)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tornado/stack_context.py\", line 277, in null_wrapper\n    return fn(*args, **kwargs)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/zmq/eventloop/zmqstream.py\", line 440, in _handle_events\n    self._handle_recv()\n  File \"/home/bekin/.local/lib/python3.6/site-packages/zmq/eventloop/zmqstream.py\", line 472, in _handle_recv\n    self._run_callback(callback, msg)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/zmq/eventloop/zmqstream.py\", line 414, in _run_callback\n    callback(*args, **kwargs)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tornado/stack_context.py\", line 277, in null_wrapper\n    return fn(*args, **kwargs)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/ipykernel/kernelbase.py\", line 281, in dispatcher\n    return self.dispatch_shell(stream, msg)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/ipykernel/kernelbase.py\", line 232, in dispatch_shell\n    handler(stream, idents, msg)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/ipykernel/kernelbase.py\", line 397, in execute_request\n    user_expressions, allow_stdin)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/ipykernel/ipkernel.py\", line 208, in do_execute\n    res = shell.run_cell(code, store_history=store_history, silent=silent)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/ipykernel/zmqshell.py\", line 533, in run_cell\n    return super(ZMQInteractiveShell, self).run_cell(*args, **kwargs)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/IPython/core/interactiveshell.py\", line 2728, in run_cell\n    interactivity=interactivity, compiler=compiler, result=result)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/IPython/core/interactiveshell.py\", line 2850, in run_ast_nodes\n    if self.run_code(code, result):\n  File \"/home/bekin/.local/lib/python3.6/site-packages/IPython/core/interactiveshell.py\", line 2910, in run_code\n    exec(code_obj, self.user_global_ns, self.user_ns)\n  File \"<ipython-input-6-1619f793995c>\", line 25, in <module>\n    strides=[1, 2, 2, 1], padding= \"SAME\") + ae_weights[\"b_visible\"]\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tensorflow/python/ops/nn_ops.py\", line 1257, in conv2d_transpose\n    name=name)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tensorflow/python/ops/gen_nn_ops.py\", line 805, in conv2d_backprop_input\n    dilations=dilations, name=name)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tensorflow/python/framework/op_def_library.py\", line 787, in _apply_op_helper\n    op_def=op_def)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tensorflow/python/framework/ops.py\", line 3271, in create_op\n    op_def=op_def)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tensorflow/python/framework/ops.py\", line 1650, in __init__\n    self._traceback = self._graph._extract_stack()  # pylint: disable=protected-access\n\nInvalidArgumentError (see above for traceback): Conv2DCustomBackpropInput: input and out_backprop must have the same batch sizeinput batch: 64outbackprop batch: 10000 batch_dim: 0\n\t [[Node: conv2d_transpose_2 = Conv2DBackpropInput[T=DT_FLOAT, data_format=\"NHWC\", dilations=[1, 1, 1, 1], padding=\"SAME\", strides=[1, 2, 2, 1], use_cudnn_on_gpu=true, _device=\"/job:localhost/replica:0/task:0/device:CPU:0\"](stack_4, Variable_22/read, Sigmoid_4)]]\n",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mInvalidArgumentError\u001b[0m                      Traceback (most recent call last)",
      "\u001b[0;32m~/.local/lib/python3.6/site-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36m_do_call\u001b[0;34m(self, fn, *args)\u001b[0m\n\u001b[1;32m   1360\u001b[0m     \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1361\u001b[0;31m       \u001b[0;32mreturn\u001b[0m \u001b[0mfn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1362\u001b[0m     \u001b[0;32mexcept\u001b[0m \u001b[0merrors\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mOpError\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.local/lib/python3.6/site-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36m_run_fn\u001b[0;34m(session, feed_dict, fetch_list, target_list, options, run_metadata)\u001b[0m\n\u001b[1;32m   1339\u001b[0m           return tf_session.TF_Run(session, options, feed_dict, fetch_list,\n\u001b[0;32m-> 1340\u001b[0;31m                                    target_list, status, run_metadata)\n\u001b[0m\u001b[1;32m   1341\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.local/lib/python3.6/site-packages/tensorflow/python/framework/errors_impl.py\u001b[0m in \u001b[0;36m__exit__\u001b[0;34m(self, type_arg, value_arg, traceback_arg)\u001b[0m\n\u001b[1;32m    515\u001b[0m             \u001b[0mcompat\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mas_text\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mc_api\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mTF_Message\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstatus\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstatus\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 516\u001b[0;31m             c_api.TF_GetCode(self.status.status))\n\u001b[0m\u001b[1;32m    517\u001b[0m     \u001b[0;31m# Delete the underlying status object from memory otherwise it stays alive\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mInvalidArgumentError\u001b[0m: Conv2DCustomBackpropInput: input and out_backprop must have the same batch sizeinput batch: 64outbackprop batch: 10000 batch_dim: 0\n\t [[Node: conv2d_transpose_2 = Conv2DBackpropInput[T=DT_FLOAT, data_format=\"NHWC\", dilations=[1, 1, 1, 1], padding=\"SAME\", strides=[1, 2, 2, 1], use_cudnn_on_gpu=true, _device=\"/job:localhost/replica:0/task:0/device:CPU:0\"](stack_4, Variable_22/read, Sigmoid_4)]]",
      "\nDuring handling of the above exception, another exception occurred:\n",
      "\u001b[0;31mInvalidArgumentError\u001b[0m                      Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-6-1619f793995c>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     40\u001b[0m     \u001b[0msess\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mconv_op\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfeed_dict\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m{\u001b[0m\u001b[0mae_input\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mx_batch\u001b[0m\u001b[0;34m}\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     41\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 42\u001b[0;31m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"\\nAccuracy\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msess\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0maccuracy\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfeed_dict\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m{\u001b[0m\u001b[0mae_input\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mmnist\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtest\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mimages\u001b[0m\u001b[0;34m}\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m~/.local/lib/python3.6/site-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36mrun\u001b[0;34m(self, fetches, feed_dict, options, run_metadata)\u001b[0m\n\u001b[1;32m    903\u001b[0m     \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    904\u001b[0m       result = self._run(None, fetches, feed_dict, options_ptr,\n\u001b[0;32m--> 905\u001b[0;31m                          run_metadata_ptr)\n\u001b[0m\u001b[1;32m    906\u001b[0m       \u001b[0;32mif\u001b[0m \u001b[0mrun_metadata\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    907\u001b[0m         \u001b[0mproto_data\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtf_session\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mTF_GetBuffer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrun_metadata_ptr\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.local/lib/python3.6/site-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36m_run\u001b[0;34m(self, handle, fetches, feed_dict, options, run_metadata)\u001b[0m\n\u001b[1;32m   1135\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mfinal_fetches\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0mfinal_targets\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mhandle\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0mfeed_dict_tensor\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1136\u001b[0m       results = self._do_run(handle, final_targets, final_fetches,\n\u001b[0;32m-> 1137\u001b[0;31m                              feed_dict_tensor, options, run_metadata)\n\u001b[0m\u001b[1;32m   1138\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1139\u001b[0m       \u001b[0mresults\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.local/lib/python3.6/site-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36m_do_run\u001b[0;34m(self, handle, target_list, fetch_list, feed_dict, options, run_metadata)\u001b[0m\n\u001b[1;32m   1353\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mhandle\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1354\u001b[0m       return self._do_call(_run_fn, self._session, feeds, fetches, targets,\n\u001b[0;32m-> 1355\u001b[0;31m                            options, run_metadata)\n\u001b[0m\u001b[1;32m   1356\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1357\u001b[0m       \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_do_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0m_prun_fn\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_session\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhandle\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfeeds\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfetches\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.local/lib/python3.6/site-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36m_do_call\u001b[0;34m(self, fn, *args)\u001b[0m\n\u001b[1;32m   1372\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0mKeyError\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1373\u001b[0m           \u001b[0;32mpass\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1374\u001b[0;31m       \u001b[0;32mraise\u001b[0m \u001b[0mtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0me\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnode_def\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mop\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmessage\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1375\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1376\u001b[0m   \u001b[0;32mdef\u001b[0m \u001b[0m_extend_graph\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mInvalidArgumentError\u001b[0m: Conv2DCustomBackpropInput: input and out_backprop must have the same batch sizeinput batch: 64outbackprop batch: 10000 batch_dim: 0\n\t [[Node: conv2d_transpose_2 = Conv2DBackpropInput[T=DT_FLOAT, data_format=\"NHWC\", dilations=[1, 1, 1, 1], padding=\"SAME\", strides=[1, 2, 2, 1], use_cudnn_on_gpu=true, _device=\"/job:localhost/replica:0/task:0/device:CPU:0\"](stack_4, Variable_22/read, Sigmoid_4)]]\n\nCaused by op 'conv2d_transpose_2', defined at:\n  File \"/usr/lib/python3.6/runpy.py\", line 193, in _run_module_as_main\n    \"__main__\", mod_spec)\n  File \"/usr/lib/python3.6/runpy.py\", line 85, in _run_code\n    exec(code, run_globals)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/ipykernel_launcher.py\", line 16, in <module>\n    app.launch_new_instance()\n  File \"/home/bekin/.local/lib/python3.6/site-packages/traitlets/config/application.py\", line 658, in launch_instance\n    app.start()\n  File \"/home/bekin/.local/lib/python3.6/site-packages/ipykernel/kernelapp.py\", line 478, in start\n    self.io_loop.start()\n  File \"/home/bekin/.local/lib/python3.6/site-packages/zmq/eventloop/ioloop.py\", line 177, in start\n    super(ZMQIOLoop, self).start()\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tornado/ioloop.py\", line 888, in start\n    handler_func(fd_obj, events)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tornado/stack_context.py\", line 277, in null_wrapper\n    return fn(*args, **kwargs)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/zmq/eventloop/zmqstream.py\", line 440, in _handle_events\n    self._handle_recv()\n  File \"/home/bekin/.local/lib/python3.6/site-packages/zmq/eventloop/zmqstream.py\", line 472, in _handle_recv\n    self._run_callback(callback, msg)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/zmq/eventloop/zmqstream.py\", line 414, in _run_callback\n    callback(*args, **kwargs)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tornado/stack_context.py\", line 277, in null_wrapper\n    return fn(*args, **kwargs)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/ipykernel/kernelbase.py\", line 281, in dispatcher\n    return self.dispatch_shell(stream, msg)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/ipykernel/kernelbase.py\", line 232, in dispatch_shell\n    handler(stream, idents, msg)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/ipykernel/kernelbase.py\", line 397, in execute_request\n    user_expressions, allow_stdin)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/ipykernel/ipkernel.py\", line 208, in do_execute\n    res = shell.run_cell(code, store_history=store_history, silent=silent)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/ipykernel/zmqshell.py\", line 533, in run_cell\n    return super(ZMQInteractiveShell, self).run_cell(*args, **kwargs)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/IPython/core/interactiveshell.py\", line 2728, in run_cell\n    interactivity=interactivity, compiler=compiler, result=result)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/IPython/core/interactiveshell.py\", line 2850, in run_ast_nodes\n    if self.run_code(code, result):\n  File \"/home/bekin/.local/lib/python3.6/site-packages/IPython/core/interactiveshell.py\", line 2910, in run_code\n    exec(code_obj, self.user_global_ns, self.user_ns)\n  File \"<ipython-input-6-1619f793995c>\", line 25, in <module>\n    strides=[1, 2, 2, 1], padding= \"SAME\") + ae_weights[\"b_visible\"]\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tensorflow/python/ops/nn_ops.py\", line 1257, in conv2d_transpose\n    name=name)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tensorflow/python/ops/gen_nn_ops.py\", line 805, in conv2d_backprop_input\n    dilations=dilations, name=name)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tensorflow/python/framework/op_def_library.py\", line 787, in _apply_op_helper\n    op_def=op_def)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tensorflow/python/framework/ops.py\", line 3271, in create_op\n    op_def=op_def)\n  File \"/home/bekin/.local/lib/python3.6/site-packages/tensorflow/python/framework/ops.py\", line 1650, in __init__\n    self._traceback = self._graph._extract_stack()  # pylint: disable=protected-access\n\nInvalidArgumentError (see above for traceback): Conv2DCustomBackpropInput: input and out_backprop must have the same batch sizeinput batch: 64outbackprop batch: 10000 batch_dim: 0\n\t [[Node: conv2d_transpose_2 = Conv2DBackpropInput[T=DT_FLOAT, data_format=\"NHWC\", dilations=[1, 1, 1, 1], padding=\"SAME\", strides=[1, 2, 2, 1], use_cudnn_on_gpu=true, _device=\"/job:localhost/replica:0/task:0/device:CPU:0\"](stack_4, Variable_22/read, Sigmoid_4)]]\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "\n",
    "mnist = input_data.read_data_sets(\"data/MNIST\", one_hot=True)\n",
    "\n",
    "batch_size, learning_rate = 64, 0.01\n",
    "\n",
    "ae_weights = {\n",
    "    \"conv\"      : tf.Variable(tf.truncated_normal([5, 5, 1, 4], stddev=0.1)),\n",
    "    \"b_hidden\"  : tf.Variable(tf.truncated_normal([4], stddev=0.1)),\n",
    "    \"deconv\"    : tf.Variable(tf.truncated_normal([5, 5, 1, 4], stddev=0.1)),\n",
    "    \"b_visible\" : tf.Variable(tf.truncated_normal([1], stddev=0.1))\n",
    "}\n",
    "\n",
    "input_shape = tf.stack([batch_size, 28, 28, 1])\n",
    "ae_input = tf.placeholder(tf.float32, [None, 784])\n",
    "images = tf.reshape(ae_input, [-1, 28, 28, 1])\n",
    "\n",
    "hidden_logits = tf.nn.conv2d(images, ae_weights[\"conv\"], \n",
    "                             strides=[1, 2, 2, 1], padding=\"SAME\") + ae_weights[\"b_hidden\"]\n",
    "hidden = tf.nn.sigmoid(hidden_logits)\n",
    "\n",
    "visible_logits = tf.nn.conv2d_transpose(hidden, ae_weights[\"deconv\"], input_shape, \n",
    "                                        strides=[1, 2, 2, 1], padding= \"SAME\") + ae_weights[\"b_visible\"]\n",
    "visible = tf.nn.sigmoid(visible_logits)\n",
    "\n",
    "optimizer = tf.train.AdagradOptimizer(learning_rate)\n",
    "conv_cost = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=visible_logits, labels=images))\n",
    "conv_op = optimizer.minimize(conv_cost)\n",
    "\n",
    "# точность\n",
    "accuracy = tf.reduce_mean(1 - tf.abs(ae_input - visible))\n",
    "\n",
    "sess = tf.Session()\n",
    "sess.run(tf.global_variables_initializer())\n",
    "\n",
    "for i in range(100000):\n",
    "    x_batch, _ = mnist.train.next_batch(batch_size)\n",
    "    sess.run(conv_op, feed_dict={ae_input: x_batch})\n",
    "\n",
    "# print(\"\\nAccuracy\", sess.run(accuracy, feed_dict={ae_input: mnist.test.images}))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Оригинал и реконструированные:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f8968648780>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import math\n",
    "\n",
    "# Reshape the array into 28 x 28 array (2-dimensional array)\n",
    "img_ae_input = tf.reshape(ae_input, [-1, 28, 28])\n",
    "img_visible = tf.reshape(visible, [-1, 28, 28])\n",
    "# img_noised_visible = tf.reshape(noised_visible, [-1, 28, 28])\n",
    "\n",
    "inp_pics, out_pics = sess.run(\n",
    "    [img_ae_input, img_visible], \n",
    "    feed_dict={ae_input: x_batch})\n",
    "\n",
    "examples_cnt = 15\n",
    "\n",
    "print(\"Оригинал и реконструированные:\")\n",
    "plt.figure(figsize=(15, 2))\n",
    "for i in range(examples_cnt):\n",
    "    plt.subplot(2, examples_cnt, i + 1)\n",
    "    plt.imshow(inp_pics[i], cmap='gray')\n",
    "\n",
    "for i in range(examples_cnt):\n",
    "    plt.subplot(2, examples_cnt, examples_cnt + i + 1)\n",
    "    plt.imshow(out_pics[i], cmap='gray')   \n",
    "plt.show()\n",
    "\n",
    "# print(\"Оригинал и восстановленные из «обрезанного» скрытого слоя:\")\n",
    "# plt.figure(figsize=(15, 2))\n",
    "# for i in range(examples_cnt):\n",
    "#     plt.subplot(2, examples_cnt, i + 1)\n",
    "#     plt.imshow(inp_pics[i], cmap='gray')\n",
    "\n",
    "# for i in range(examples_cnt):\n",
    "#     plt.subplot(2, examples_cnt, examples_cnt + i + 1)\n",
    "#     plt.imshow(noised_out_pics[i], cmap='gray')\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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}