{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import faiss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "index = faiss.index_factory(128, \"IVF52428_HNSW32,PQ64\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# train \n",
    "xt = faiss.rand((100000 ,128))\n",
    "index.train(xt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# initial add\n",
    "x = faiss.rand((130000, 128))\n",
    "index.add(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "index size 280000, mem usage 547.32 MiB\r"
     ]
    }
   ],
   "source": [
    "# add 150k by batches of 100, monitor mem usage\n",
    "log = []\n",
    "for na in range(0, 150000, 100): \n",
    "    x = faiss.rand((100, 128), seed=na * 12345)\n",
    "    index.add(x)\n",
    "    mem = faiss.get_mem_usage_kb()\n",
    "    print('index size %d, mem usage %.2f MiB' % (index.ntotal, mem / 1024.), end='\\r', flush=True)\n",
    "    log.append((index.ntotal, mem))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "87.8221747831888"
      ]
     },
     "execution_count": 13,
     "metadata": {
      "bento_obj_id": "140247899674688"
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# added bytes per vector\n",
    "\n",
    "n0, mem0 = log[0]\n",
    "n1, mem1 = log[-1]\n",
    "\n",
    "(mem1 - mem0) * 1024 / (n1 - n0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "bento_stylesheets": {
   "bento/extensions/flow/main.css": true,
   "bento/extensions/kernel_selector/main.css": true,
   "bento/extensions/kernel_ui/main.css": true,
   "bento/extensions/new_kernel/main.css": true,
   "bento/extensions/system_usage/main.css": true,
   "bento/extensions/theme/main.css": true
  },
  "kernelspec": {
   "display_name": "faiss",
   "language": "python",
   "name": "bento_kernel_faiss"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3rc1+"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}

   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import faiss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "index = faiss.index_factory(128, \"IVF52428_HNSW32,PQ64\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# train \n",
    "xt = faiss.rand((100000 ,128))\n",
    "index.train(xt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# initial add\n",
    "x = faiss.rand((130000, 128))\n",
    "index.add(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "index size 280000, mem usage 547.32 MiB\r"
     ]
    }
   ],
   "source": [
    "# add 150k by batches of 100, monitor mem usage\n",
    "log = []\n",
    "for na in range(0, 150000, 100): \n",
    "    x = faiss.rand((100, 128), seed=na * 12345)\n",
    "    index.add(x)\n",
    "    mem = faiss.get_mem_usage_kb()\n",
    "    print('index size %d, mem usage %.2f MiB' % (index.ntotal, mem / 1024.), end='\\r', flush=True)\n",
    "    log.append((index.ntotal, mem))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "87.8221747831888"
      ]
     },
     "execution_count": 13,
     "metadata": {
      "bento_obj_id": "140247899674688"
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# added bytes per vector\n",
    "\n",
    "n0, mem0 = log[0]\n",
    "n1, mem1 = log[-1]\n",
    "\n",
    "(mem1 - mem0) * 1024 / (n1 - n0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "bento_stylesheets": {
   "bento/extensions/flow/main.css": true,
   "bento/extensions/kernel_selector/main.css": true,
   "bento/extensions/kernel_ui/main.css": true,
   "bento/extensions/new_kernel/main.css": true,
   "bento/extensions/system_usage/main.css": true,
   "bento/extensions/theme/main.css": true
  },
  "kernelspec": {
   "display_name": "faiss",
   "language": "python",
   "name": "bento_kernel_faiss"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3rc1+"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}