model_2-2_7.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ubuntu/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n",
      "/home/ubuntu/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/matplotlib/__init__.py:962: UserWarning: Duplicate key in file \"/home/ubuntu/.config/matplotlib/matplotlibrc\", line #2\n",
      "  (fname, cnt))\n",
      "/home/ubuntu/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/matplotlib/__init__.py:962: UserWarning: Duplicate key in file \"/home/ubuntu/.config/matplotlib/matplotlibrc\", line #3\n",
      "  (fname, cnt))\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import model_generator as mg\n",
    "inIter = '6'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = mg.Model([300, 400, 40, 1], 'Model_2-2', \n",
    "                 dataFilename = 'Model_2-2/T_'+inIter+'.h5', \n",
    "                 paramFileName = 'Model_2-2/params_Model_2-2_'+inIter+'.pkl', r_discount=.95, c_discount=1)\n",
    "mg.saveHP(model)\n",
    "model.setData()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "num minibatches:1506\n",
      "\tEpoch time:21 seconds\n",
      "Cost after epoch 0: 0.776961\n",
      "\tDev error (no dropout): 0.749730\n",
      "Cost after epoch 1: 0.765826\n",
      "\tDev error (no dropout): 0.743924\n",
      "Cost after epoch 2: 0.761409\n",
      "\tDev error (no dropout): 0.739331\n",
      "Cost after epoch 3: 0.757259\n",
      "\tDev error (no dropout): 0.734223\n",
      "Cost after epoch 4: 0.754650\n",
      "\tDev error (no dropout): 0.731946\n",
      "Cost after epoch 5: 0.750828\n",
      "\tDev error (no dropout): 0.729515\n",
      "Cost after epoch 6: 0.749581\n",
      "\tDev error (no dropout): 0.728077\n",
      "Cost after epoch 7: 0.747117\n",
      "\tDev error (no dropout): 0.727588\n",
      "Cost after epoch 8: 0.744910\n",
      "\tDev error (no dropout): 0.729817\n",
      "Cost after epoch 9: 0.743337\n",
      "\tDev error (no dropout): 0.727133\n",
      "\tAve epoch time: 21.506042\n",
      "\t100 epochs will take 0:35:50.604184\n",
      "Cost after epoch 10: 0.741662\n",
      "\tDev error (no dropout): 0.724585\n",
      "Cost after epoch 20: 0.730361\n",
      "\tDev error (no dropout): 0.719988\n",
      "Parameters have been trained!\n",
      "Train cost (No dropout): 0.7027445\n",
      "Test cost (No dropout): 0.71972775\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f509e9406a0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.TrainModel(num_epochs=30, minibatch_size=2048, learning_rate=0.0001)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "VS Greedy: 62.800%\tVS Random: 94.700%\n",
      "Standard metric: 197.16\n"
     ]
    }
   ],
   "source": [
    "model.TestModel(1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "VS Greedy: 62.475%\tVS Random: 94.050%\n",
      "Standard metric: 223.27\n"
     ]
    }
   ],
   "source": [
    "model.TestModel(20000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "os.system('git add .')\n",
    "os.system('git commit -m \"Model_2-2_'+str(model.modelIter)+' trained\"')\n",
    "os.system('git pull')\n",
    "os.system('git push origin master')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import zsyMultigame as multi\n",
    "import os\n",
    "outIter = ''\n",
    "multi.simulate('Model_2-2/params_Model_2-2_'+outIter+'.pkl', os.path.join('Model_2-2','T_'+outIter+'.h5'),\n",
    "               numGames=100000, poolsize = 20, exploration_prob = 0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "os.system('git add .')\n",
    "os.system('git commit -m \"Model_2-2_'+outIter+' simulated\"')\n",
    "os.system('git pull')\n",
    "os.system('git push origin master')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Environment (conda_tensorflow_p36)",
   "language": "python",
   "name": "conda_tensorflow_p36"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}