Initialize RNN building block notebook

RNN_building_blocks.ipynb

@@ -0,0 +1,60 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "collapsed": true,
+    "pycharm": {
+     "name": "#%% md\n"
+    }
+   },
+   "source": [
+    "# RNN Building Blocks: injecting recursivity to supervised learner"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "outputs": [],
+   "source": [
+    "import torch\n",
+    "from torchvision import datasets, transforms\n",
+    "import torch.nn.functional as F\n",
+    "from torch import nn\n",
+    "torch.manual_seed(0)\n",
+    "torch.backends.cudnn.deterministic = True\n",
+    "torch.backends.cudnn.benchmark = False\n",
+    "\n",
+    "import numpy as np\n",
+    "np.random.seed(0)"
+   ],
+   "metadata": {
+    "collapsed": false,
+    "pycharm": {
+     "name": "#%%\n"
+    }
+   }
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "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
+}