{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-01-18T08:48:06.489049Z",
     "start_time": "2018-01-18T08:48:06.486256Z"
    }
   },
   "outputs": [],
   "source": [
    "import os\n",
    "os.environ['CUDA_VISIBLE_DEVICES']=\"\"\n",
    "os.environ[\"PYTHONPATH\"]='.'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-01-18T08:48:06.842630Z",
     "start_time": "2018-01-18T08:48:06.490286Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Populating the interactive namespace from numpy and matplotlib\n"
     ]
    }
   ],
   "source": [
    "%pylab --no-import-all inline\n",
    "%reload_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-01-18T08:48:06.881370Z",
     "start_time": "2018-01-18T08:48:06.844048Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['ddpg/train.py',\n",
       " '--logdir',\n",
       " './outputs/logs_ddpg',\n",
       " '--num-threads',\n",
       " '1',\n",
       " '--ddpg-wrapper',\n",
       " '--skip-frames',\n",
       " '5',\n",
       " '--fail-reward',\n",
       " '-0.2',\n",
       " '--reward-scale',\n",
       " '1',\n",
       " '--flip-state-action',\n",
       " '--actor-layers',\n",
       " '64-64',\n",
       " '--actor-layer-norm',\n",
       " '--actor-parameters-noise',\n",
       " '--actor-lr',\n",
       " '0.001',\n",
       " '--actor-lr-end',\n",
       " '0.00001',\n",
       " '--critic-layers',\n",
       " '64-32',\n",
       " '--critic-layer-norm',\n",
       " '--critic-lr',\n",
       " '0.002',\n",
       " '--critic-lr-end',\n",
       " '0.00001',\n",
       " '--initial-epsilon',\n",
       " '0.5',\n",
       " '--final-epsilon',\n",
       " '0.001',\n",
       " '--tau',\n",
       " '0.0001']"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "os.sys.argv=\"ddpg/train.py --logdir ./outputs/logs_ddpg --num-threads 1 --ddpg-wrapper --skip-frames 5 --fail-reward -0.2 --reward-scale 1 --flip-state-action --actor-layers 64-64 --actor-layer-norm --actor-parameters-noise --actor-lr 0.001 --actor-lr-end 0.00001 --critic-layers 64-32 --critic-layer-norm --critic-lr 0.002 --critic-lr-end 0.00001 --initial-epsilon 0.5 --final-epsilon 0.001 --tau 0.0001\".split(\" \")\n",
    "os.sys.argv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-01-18T08:49:02.648421Z",
     "start_time": "2018-01-18T08:49:02.573206Z"
    }
   },
   "outputs": [],
   "source": [
    "from ddpg.train import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "start_time": "2018-01-18T08:49:03.555Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[2018-01-18 16:49:08,700] Making new env: Pendulum-v0\n",
      "[2018-01-18 16:49:08,777] Making new env: Pendulum-v0\n",
      "[2018-01-18 16:49:08,780] Making new env: Pendulum-v0\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Base (\n",
      "  (feature_net): LinearNet (\n",
      "    (net): Sequential (\n",
      "      (linear_0): NoisyLinear (3 -> 64)\n",
      "      (layer_norm_0): LayerNorm (\n",
      "      )\n",
      "      (act_0): ReLU ()\n",
      "      (linear_1): NoisyLinear (64 -> 64)\n",
      "      (layer_norm_1): LayerNorm (\n",
      "      )\n",
      "      (act_1): ReLU ()\n",
      "    )\n",
      "  )\n",
      ")\n",
      "ActorHead (\n",
      "  (base): Base (\n",
      "    (feature_net): LinearNet (\n",
      "      (net): Sequential (\n",
      "        (linear_0): NoisyLinear (3 -> 64)\n",
      "        (layer_norm_0): LayerNorm (\n",
      "        )\n",
      "        (act_0): ReLU ()\n",
      "        (linear_1): NoisyLinear (64 -> 64)\n",
      "        (layer_norm_1): LayerNorm (\n",
      "        )\n",
      "        (act_1): ReLU ()\n",
      "      )\n",
      "    )\n",
      "  )\n",
      "  (policy_net): LinearNet (\n",
      "    (net): Sequential (\n",
      "      (linear_0): Linear (64 -> 1)\n",
      "      (act_0): Tanh ()\n",
      "    )\n",
      "  )\n",
      ")\n",
      "CriticHead (\n",
      "  (base): Base (\n",
      "    (feature_net): LinearNet (\n",
      "      (net): Sequential (\n",
      "        (linear_0): NoisyLinear (3 -> 64)\n",
      "        (layer_norm_0): LayerNorm (\n",
      "        )\n",
      "        (act_0): ReLU ()\n",
      "        (linear_1): NoisyLinear (64 -> 64)\n",
      "        (layer_norm_1): LayerNorm (\n",
      "        )\n",
      "        (act_1): ReLU ()\n",
      "      )\n",
      "    )\n",
      "  )\n",
      "  (value_net): Linear (64 -> 1)\n",
      ")\n",
      "DynamicsHead (\n",
      "  (base): Base (\n",
      "    (feature_net): LinearNet (\n",
      "      (net): Sequential (\n",
      "        (linear_0): NoisyLinear (3 -> 64)\n",
      "        (layer_norm_0): LayerNorm (\n",
      "        )\n",
      "        (act_0): ReLU ()\n",
      "        (linear_1): NoisyLinear (64 -> 64)\n",
      "        (layer_norm_1): LayerNorm (\n",
      "        )\n",
      "        (act_1): ReLU ()\n",
      "      )\n",
      "    )\n",
      "  )\n",
      "  (value_net): Linear (65 -> 3)\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "os.environ['OMP_NUM_THREADS'] = '1'\n",
    "torch.set_num_threads(1)\n",
    "args = parse_args()\n",
    "train(args,\n",
    "      create_model,\n",
    "      create_act_update_fns,\n",
    "      train_multi_thread,\n",
    "      train_single_thread,\n",
    "      play_single_thread)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "start_time": "2018-01-18T08:39:00.927Z"
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "start_time": "2018-01-18T08:39:00.929Z"
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "jupyter3",
   "language": "python",
   "name": "jupyter3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.0"
  },
  "toc": {
   "colors": {
    "hover_highlight": "#DAA520",
    "navigate_num": "#000000",
    "navigate_text": "#333333",
    "running_highlight": "#FF0000",
    "selected_highlight": "#FFD700",
    "sidebar_border": "#EEEEEE",
    "wrapper_background": "#FFFFFF"
   },
   "moveMenuLeft": true,
   "nav_menu": {
    "height": "12px",
    "width": "252px"
   },
   "navigate_menu": true,
   "number_sections": true,
   "sideBar": true,
   "threshold": 4,
   "toc_cell": false,
   "toc_section_display": "block",
   "toc_window_display": false,
   "widenNotebook": false
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}