mogym/Experiments.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "089df536",
   "metadata": {},
   "source": [
    "# Notebook to Reproduce the Results Demonstrated in the Paper\n",
    "\n",
    "In the following, we give step by step instructions on how to reproduce the results we presented in the paper.\n",
    "\n",
    "We show (a) how to train an NN (Sect. 2) using MoGym's API and (b) how to asses the quality of a trained or arbitrary decision agent (Sect. 3 and 4).\n",
    "\n",
    "This artifact comprises four experiments: *racetrack*, *cdrive*, *elevators* and *firewire*.\n",
    "\n",
    "For each of the experiments there are two files in the `mogym` directory (same directory as this notebook): `learn_EXPERIMENT.py` and `learn_EXPERIMENT.hermes`.\n",
    "\n",
    "- The `learn_EXPERIMENT.py` file contains Python code using the MoGym API (in particular the function `create_generic_env`) to construct an Open AI Gym compatible environment. The environment is then used to construct an `Agent` for training (using the `train` method on the constructed `Agent`). The `Agent` we are using here is defined in the [`dqn.py`](./dqn.py) file and implements *Deep Q-Learning*.\n",
    "- The `learn_EXPERIMENT.hermes` is a configuration file for the reinforcement learning toolkit [RLMate](https://pypi.org/project/rlmate/). This file specifies various parameters such as the structure of the *Neural Network* (NN) that should be used and various other parameters for the experiment. Note that reinforcement learning requires tweaking these so called *hyperparameters*. We have chosen them such that learning actually works well with the respective models. The general structure of the NNs we are using is defined in [`fcn.py`](./fcn.py).\n",
    "\n",
    "In the following, there are *code cells* for executing *Hermes*—a command line interface provided by RLMate which invokes the `learn_EXPERIMENT.py` scripts with the parameters specified in the respective `learn_EXPERIMENT.hermes` file and stores the *learning artifacts* (i.e., the trained NN, learning curves, et cetera) in the [`experiments`](../experiments/) directory (at the root of the artifact).\n",
    "\n",
    "The same [`experiments`](../experiments/) directory also contains the neural networks we used for the results in the paper as well as other learning artifacts. When plotting the learning curves these results will be used. To instead plot the learning curves of your own experiments, you need to change the path with which `storage.Experiment.load` is invoked accordingly.\n",
    "\n",
    "To execute a code cell, press `SHIFT + ENTER` (or use the run button in the toolbar at the top). A number in brackets on the left indicates that the cell is done running. A `*` indicates that the cell is still executing. At the beginning of each section (and of this notebook) there is a cell containing necessary imports and other setup code. Please make sure to execute all the cells in the order they are defined, otherwise imports and other things may be missing. If you want to save time, you can skip the execution of the `hermes` cells for learning.\n",
    "\n",
    "We provide the output of an execution of the cells on our machine for your convenience.\n",
    "\n",
    "Here is such a cell. Please execute it:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "d5fb711f-9154-406c-8e29-bbff00ff426d",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pathlib\n",
    "import rlmate.storage as storage\n",
    "import rlmate.plot as hplt\n",
    "\n",
    "# disable some warnings not related to our experiments\n",
    "import warnings\n",
    "warnings.simplefilter(\"ignore\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dc8fdcf9",
   "metadata": {},
   "source": [
    "# 1. Train an NN (Sect. 2 of the paper)\n",
    "\n",
    "The following commands can be executed in order to learn a neural network for the respective models.\n",
    "\n",
    "## 1.1. Racetrack\n",
    "\n",
    "For *Racetrack* we use the Python Package [racetrack](https://pypi.org/project/racetrack/), which provides a configurable JANI model for the racetrack benchmark. \n",
    "\n",
    "By executing the following cell, our racetrack experiments are reproduced. \n",
    "\n",
    "Note that learning an NN for racetrack takes a considerable amount of RAM and time. Depending on your machine, this might not work for Racetrack. We recommend to use at least 24 GB of RAM. You can skip the next command and instead execute the learning commands for the other models, which are less resource consuming.\n",
    "\n",
    "Executing the `hermes` command took around 3 1/4 hours on our machine.\n",
    "Similar experiments for the `cdrive.2` model described below will take less time and resources."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "d4f5263e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[2022-05-06 05:27:07,960] Launching Hermes with experiments path set to ./\n",
      "[2022-05-06 05:27:07,960] Executing from hermes execution file learn_racetrack.hermes\n",
      "[2022-05-06 05:27:07,965] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 05:27:07,966] Creating job directory /workspace/mogym/../experiments/learn_racetrack/racetrack/2022_05_06_052707962018_0_00000\n",
      "[2022-05-06 05:27:07,967] Launching job Nr. 0: \"python learn_racetrack.py 2022_05_06_052707962018_0_00000 -nnw 15 64 64 9 -nnf fcn -ne 20000 -es 0.05 -ee 0.05 -ed 0.999 -pr 1  -nr -0.2 -pef 1000 -ne 20000 -s 1\".\n",
      "[2022-05-06 05:27:07,968] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 06:07:37,388] Sucessfully finished job Nr. 0 2022_05_06_052707962018_0_00000 in 2429.42 seconds\n",
      "[2022-05-06 06:07:37,403] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 06:07:37,404] Creating job directory /workspace/mogym/../experiments/learn_racetrack/racetrack/2022_05_06_060737399494_1_00000\n",
      "[2022-05-06 06:07:37,405] Launching job Nr. 1: \"python learn_racetrack.py 2022_05_06_060737399494_1_00000 -nnw 15 64 64 9 -nnf fcn -ne 20000 -es 0.05 -ee 0.05 -ed 0.999 -pr 1  -nr -0.2 -pef 1000 -ne 20000 -s 2\".\n",
      "[2022-05-06 06:07:37,405] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 06:35:21,336] Sucessfully finished job Nr. 1 2022_05_06_060737399494_1_00000 in 1663.93 seconds\n",
      "[2022-05-06 06:35:21,348] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 06:35:21,348] Creating job directory /workspace/mogym/../experiments/learn_racetrack/racetrack/2022_05_06_063521344598_2_00000\n",
      "[2022-05-06 06:35:21,349] Launching job Nr. 2: \"python learn_racetrack.py 2022_05_06_063521344598_2_00000 -nnw 15 64 64 9 -nnf fcn -ne 20000 -es 0.05 -ee 0.05 -ed 0.999 -pr 1  -nr -0.2 -pef 1000 -ne 20000 -s 3\".\n",
      "[2022-05-06 06:35:21,349] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 07:08:23,330] Sucessfully finished job Nr. 2 2022_05_06_063521344598_2_00000 in 1981.98 seconds\n",
      "[2022-05-06 07:08:23,342] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 07:08:23,343] Creating job directory /workspace/mogym/../experiments/learn_racetrack/racetrack/2022_05_06_070823339133_3_00000\n",
      "[2022-05-06 07:08:23,344] Launching job Nr. 3: \"python learn_racetrack.py 2022_05_06_070823339133_3_00000 -nnw 15 64 64 9 -nnf fcn -ne 20000 -es 0.05 -ee 0.05 -ed 0.999 -pr 1  -nr -0.2 -pef 1000 -ne 20000 -s 4\".\n",
      "[2022-05-06 07:08:23,344] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 07:38:49,331] Sucessfully finished job Nr. 3 2022_05_06_070823339133_3_00000 in 1825.99 seconds\n",
      "[2022-05-06 07:38:49,343] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 07:38:49,344] Creating job directory /workspace/mogym/../experiments/learn_racetrack/racetrack/2022_05_06_073849339995_4_00000\n",
      "[2022-05-06 07:38:49,345] Launching job Nr. 4: \"python learn_racetrack.py 2022_05_06_073849339995_4_00000 -nnw 15 64 64 9 -nnf fcn -ne 20000 -es 0.05 -ee 0.05 -ed 0.999 -pr 1  -nr -0.2 -pef 1000 -ne 20000 -s 5\".\n",
      "[2022-05-06 07:38:49,346] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 08:07:46,299] Sucessfully finished job Nr. 4 2022_05_06_073849339995_4_00000 in 1736.95 seconds\n",
      "[2022-05-06 08:07:46,311] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 08:07:46,312] Creating job directory /workspace/mogym/../experiments/learn_racetrack/racetrack/2022_05_06_080746308538_5_00000\n",
      "[2022-05-06 08:07:46,313] Launching job Nr. 5: \"python learn_racetrack.py 2022_05_06_080746308538_5_00000 -nnw 15 64 64 9 -nnf fcn -ne 20000 -es 0.05 -ee 0.05 -ed 0.999 -pr 1  -nr -0.2 -pef 1000 -ne 20000 -s 6\".\n",
      "[2022-05-06 08:07:46,313] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 08:36:29,291] Sucessfully finished job Nr. 5 2022_05_06_080746308538_5_00000 in 1722.98 seconds\n",
      "CPU times: user 4min 50s, sys: 1min 18s, total: 6min 8s\n",
      "Wall time: 3h 9min 23s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "!hermes exec -f learn_racetrack.hermes -n 'racetrack' --root '../'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7a2ab631",
   "metadata": {},
   "source": [
    "The [`learn_racetrack.hermes`](./learn_racetrack.hermes) file specifies how the job shall be processed: We make use of the neural network defined in [`fcn.py`](./fcn.py) (`-nnf`) using the following structure `15 64 64 9` (`-nnw`) (15 input neurons, 9 output neurons, and two hidden layers of size 64 in between). The exploration coefficient epsilon is in the beginning set to 1 (`-es`) and is exponentially decayed until the value 0.001 (`-ee`) is reached. The training takes 8000 episodes (`-ne`). Further, this file specifies that the script [`learn_racetrack.py`](./learn_racetrack.py) should be used for training. In the latter one can find\n",
    "- the initialization of the racetrack environment with the MoGym package and its API,\n",
    "- the initialization of the DQN agent, and\n",
    "- the actual start of the training.\n",
    "\n",
    "\n",
    "All of the following experiments work analogously, which is why we don't provide another detailed description for them.\n",
    "\n",
    "The following code produces and displays the training curve (which is a part of Fig. 3 b) of the paper):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "a3d310b7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<module 'matplotlib.pyplot' from '/home/vscode/.local/lib/python3.9/site-packages/matplotlib/pyplot.py'>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 864x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "experiment = storage.Experiment.load('../experiments/learn_racetrack/racetrack2/2022_01_20_210439836184_5_fbdf8')\n",
    "scores = experiment.scores()\n",
    "hplt.plot_training_progress(scores)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "960c2789",
   "metadata": {},
   "source": [
    "## 1.2. cdrive\n",
    "\n",
    "For *cdrive* (using the `cdrive.2` QVBS model) execute:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "40d58e6a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[2022-05-06 08:36:33,211] Launching Hermes with experiments path set to ./\n",
      "[2022-05-06 08:36:33,211] Executing from hermes execution file learn_cdrive.hermes\n",
      "[2022-05-06 08:36:33,215] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 08:36:33,217] Creating job directory /workspace/mogym/../experiments/learn_cdrive/cdrive/2022_05_06_083633213025_0_00000\n",
      "[2022-05-06 08:36:33,219] Launching job Nr. 0: \"python learn_cdrive.py 2022_05_06_083633213025_0_00000 -nnw 7 64 64 45 -nnf fcn -ed 0.999 -ee 0.05 -es 0.05 -pef 1000 -pr 1 -nr -0.2 -g 0.99 -ne 10000 -s 1\".\n",
      "[2022-05-06 08:36:33,219] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 08:38:05,598] Sucessfully finished job Nr. 0 2022_05_06_083633213025_0_00000 in 92.38 seconds\n",
      "[2022-05-06 08:38:05,602] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 08:38:05,602] Creating job directory /workspace/mogym/../experiments/learn_cdrive/cdrive/2022_05_06_083805600990_1_00000\n",
      "[2022-05-06 08:38:05,603] Launching job Nr. 1: \"python learn_cdrive.py 2022_05_06_083805600990_1_00000 -nnw 7 64 64 45 -nnf fcn -ed 0.999 -ee 0.05 -es 0.05 -pef 1000 -pr 1 -nr -0.2 -g 0.99 -ne 10000 -s 2\".\n",
      "[2022-05-06 08:38:05,603] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 08:39:36,833] Sucessfully finished job Nr. 1 2022_05_06_083805600990_1_00000 in 91.23 seconds\n",
      "[2022-05-06 08:39:36,842] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 08:39:36,842] Creating job directory /workspace/mogym/../experiments/learn_cdrive/cdrive/2022_05_06_083936839335_2_00000\n",
      "[2022-05-06 08:39:36,844] Launching job Nr. 2: \"python learn_cdrive.py 2022_05_06_083936839335_2_00000 -nnw 7 64 64 45 -nnf fcn -ed 0.999 -ee 0.05 -es 0.05 -pef 1000 -pr 1 -nr -0.2 -g 0.99 -ne 10000 -s 3\".\n",
      "[2022-05-06 08:39:36,844] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 08:40:00,354] Sucessfully finished job Nr. 2 2022_05_06_083936839335_2_00000 in 23.51 seconds\n",
      "[2022-05-06 08:40:00,362] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 08:40:00,363] Creating job directory /workspace/mogym/../experiments/learn_cdrive/cdrive/2022_05_06_084000359693_3_00000\n",
      "[2022-05-06 08:40:00,364] Launching job Nr. 3: \"python learn_cdrive.py 2022_05_06_084000359693_3_00000 -nnw 7 64 64 45 -nnf fcn -ed 0.999 -ee 0.05 -es 0.05 -pef 1000 -pr 1 -nr -0.2 -g 0.99 -ne 10000 -s 4\".\n",
      "[2022-05-06 08:40:00,364] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 08:41:21,653] Sucessfully finished job Nr. 3 2022_05_06_084000359693_3_00000 in 81.29 seconds\n",
      "[2022-05-06 08:41:21,661] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 08:41:21,662] Creating job directory /workspace/mogym/../experiments/learn_cdrive/cdrive/2022_05_06_084121659191_4_00000\n",
      "[2022-05-06 08:41:21,664] Launching job Nr. 4: \"python learn_cdrive.py 2022_05_06_084121659191_4_00000 -nnw 7 64 64 45 -nnf fcn -ed 0.999 -ee 0.05 -es 0.05 -pef 1000 -pr 1 -nr -0.2 -g 0.99 -ne 10000 -s 5\".\n",
      "[2022-05-06 08:41:21,664] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 08:42:05,619] Sucessfully finished job Nr. 4 2022_05_06_084121659191_4_00000 in 43.96 seconds\n",
      "[2022-05-06 08:42:05,628] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 08:42:05,628] Creating job directory /workspace/mogym/../experiments/learn_cdrive/cdrive/2022_05_06_084205625336_5_00000\n",
      "[2022-05-06 08:42:05,630] Launching job Nr. 5: \"python learn_cdrive.py 2022_05_06_084205625336_5_00000 -nnw 7 64 64 45 -nnf fcn -ed 0.999 -ee 0.05 -es 0.05 -pef 1000 -pr 1 -nr -0.2 -g 0.99 -ne 10000 -s 6\".\n",
      "[2022-05-06 08:42:05,630] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 08:43:47,752] Sucessfully finished job Nr. 5 2022_05_06_084205625336_5_00000 in 102.12 seconds\n",
      "CPU times: user 8.26 s, sys: 2.2 s, total: 10.5 s\n",
      "Wall time: 7min 17s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "!hermes exec -f learn_cdrive.hermes -n 'cdrive'  --root '../'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "591dc8be",
   "metadata": {},
   "source": [
    "This took about 7 minutes on our machine to complete.\n",
    "\n",
    "In contrast to *Racetrack*, we load an actual JANI model here using the Momba function `jani.load` (see [learn_cdrive.py](./learn_cdrive.py)).\n",
    "\n",
    "The following code produces and displays the training curve (which is a part of Fig. 3 a) of the paper):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "04ae9aa7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<module 'matplotlib.pyplot' from '/home/vscode/.local/lib/python3.9/site-packages/matplotlib/pyplot.py'>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 864x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "experiment = storage.Experiment.load('../experiments/learn_cdrive/cdrive/2022_01_20_133149959763_0_664e5/')\n",
    "scores = experiment.scores()\n",
    "hplt.plot_training_progress(scores)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dd77dbc2",
   "metadata": {},
   "source": [
    "### 1.3. elevators\n",
    "\n",
    "Executing the learning script for *elevators* took around 21 minutes on our machine. Three different NNs are trained:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "82498991",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[2022-05-06 08:43:50,942] Launching Hermes with experiments path set to ./\n",
      "[2022-05-06 08:43:50,943] Executing from hermes execution file learn_elevators.hermes\n",
      "[2022-05-06 08:43:50,946] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 08:43:50,947] Creating job directory /workspace/mogym/../experiments/learn_elevators/elevators/2022_05_06_084350944296_0_00000\n",
      "[2022-05-06 08:43:50,949] Launching job Nr. 0: \"python learn_elevators.py 2022_05_06_084350944296_0_00000 -nnw 22 64 64 41 -nnf fcn -ed 0.99 -ne 8000 -es 1 -ee 0.001 -ed 0.99 -s 0\".\n",
      "[2022-05-06 08:43:50,949] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 08:51:02,769] Sucessfully finished job Nr. 0 2022_05_06_084350944296_0_00000 in 431.82 seconds\n",
      "[2022-05-06 08:51:02,776] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 08:51:02,777] Creating job directory /workspace/mogym/../experiments/learn_elevators/elevators/2022_05_06_085102772990_1_00000\n",
      "[2022-05-06 08:51:02,778] Launching job Nr. 1: \"python learn_elevators.py 2022_05_06_085102772990_1_00000 -nnw 22 64 64 41 -nnf fcn -ed 0.99 -ne 8000 -es 1 -ee 0.001 -ed 0.99 -s 1\".\n",
      "[2022-05-06 08:51:02,778] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 08:58:08,556] Sucessfully finished job Nr. 1 2022_05_06_085102772990_1_00000 in 425.78 seconds\n",
      "[2022-05-06 08:58:08,558] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 08:58:08,559] Creating job directory /workspace/mogym/../experiments/learn_elevators/elevators/2022_05_06_085808557842_2_00000\n",
      "[2022-05-06 08:58:08,559] Launching job Nr. 2: \"python learn_elevators.py 2022_05_06_085808557842_2_00000 -nnw 22 64 64 41 -nnf fcn -ed 0.99 -ne 8000 -es 1 -ee 0.001 -ed 0.99 -s 2\".\n",
      "[2022-05-06 08:58:08,559] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 09:05:36,732] Sucessfully finished job Nr. 2 2022_05_06_085808557842_2_00000 in 448.17 seconds\n",
      "CPU times: user 23 s, sys: 6.55 s, total: 29.5 s\n",
      "Wall time: 21min 48s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "!hermes exec -f learn_elevators.hermes -n 'elevators' --root '../'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6d97197b",
   "metadata": {},
   "source": [
    "The following code produces and displays a training curve. You can also switch the experiment number (`2021_11_03_165704354043_c5bb2`) to display another of the experiments. Have a look in the respective experiments directory for additional experiments."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "f4d94098",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:root:Failed to load experiment start time.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<module 'matplotlib.pyplot' from '/home/vscode/.local/lib/python3.9/site-packages/matplotlib/pyplot.py'>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 864x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "experiment = storage.Experiment.load('../experiments/learn_elevators/elevators/2021_11_03_165704354043_c5bb2/')\n",
    "scores = experiment.scores()\n",
    "hplt.plot_training_progress(scores)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5376e98b",
   "metadata": {},
   "source": [
    "### 1.4. firewire_dl\n",
    "\n",
    "Executing the learning script for *firewire* took around 20 minutes on our machine. Again 3 NNs are trained:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "52f6e394",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[2022-05-06 09:05:39,850] Launching Hermes with experiments path set to ./\n",
      "[2022-05-06 09:05:39,850] Executing from hermes execution file learn_firewire.hermes\n",
      "[2022-05-06 09:05:39,853] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 09:05:39,855] Creating job directory /workspace/mogym/../experiments/learn_firewire/firewire/2022_05_06_090539851857_0_00000\n",
      "[2022-05-06 09:05:39,857] Launching job Nr. 0: \"python learn_firewire.py 2022_05_06_090539851857_0_00000 -nnf fcn -nnw 3 64 64 24 -ne 8000 -es 0.05 -ee 0.05 -ed 0.99 -s 0\".\n",
      "[2022-05-06 09:05:39,857] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 09:11:59,358] Sucessfully finished job Nr. 0 2022_05_06_090539851857_0_00000 in 379.50 seconds\n",
      "[2022-05-06 09:11:59,365] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 09:11:59,365] Creating job directory /workspace/mogym/../experiments/learn_firewire/firewire/2022_05_06_091159362023_1_00000\n",
      "[2022-05-06 09:11:59,367] Launching job Nr. 1: \"python learn_firewire.py 2022_05_06_091159362023_1_00000 -nnf fcn -nnw 3 64 64 24 -ne 8000 -es 0.05 -ee 0.05 -ed 0.99 -s 1\".\n",
      "[2022-05-06 09:11:59,367] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 09:18:32,149] Sucessfully finished job Nr. 1 2022_05_06_091159362023_1_00000 in 392.78 seconds\n",
      "[2022-05-06 09:18:32,157] Unable to retrieve commit hash for repo head\n",
      "[2022-05-06 09:18:32,158] Creating job directory /workspace/mogym/../experiments/learn_firewire/firewire/2022_05_06_091832153929_2_00000\n",
      "[2022-05-06 09:18:32,159] Launching job Nr. 2: \"python learn_firewire.py 2022_05_06_091832153929_2_00000 -nnf fcn -nnw 3 64 64 24 -ne 8000 -es 0.05 -ee 0.05 -ed 0.99 -s 2\".\n",
      "[2022-05-06 09:18:32,159] STDOUT and STDERR will be written to log files.\n",
      "[2022-05-06 09:25:03,965] Sucessfully finished job Nr. 2 2022_05_06_091832153929_2_00000 in 391.81 seconds\n",
      "CPU times: user 22.2 s, sys: 5.51 s, total: 27.8 s\n",
      "Wall time: 19min 26s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "!hermes exec -f learn_firewire.hermes -n 'firewire' --root '../'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "99822baf",
   "metadata": {},
   "source": [
    "The following code produces and displays a training curve. You can also switch the experiment number to display another of the experiments. Have a look in the respective experiments directory for additional experiments."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "33df3054",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:root:Failed to load experiment start time.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<module 'matplotlib.pyplot' from '/home/vscode/.local/lib/python3.9/site-packages/matplotlib/pyplot.py'>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAt8AAALJCAYAAABshU9CAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAACM+ElEQVR4nO3dd/xT1f3H8fdhfvkCIgKKC1AQcIEKrjpQ62hduPdExF21VqtVK2qtq6JWrQvFUUVtRbTqr4ID3AMcYEUEBFFUZO/N+f1xcslNvkm+GXfk+83r+XjkcZObO05ubm4+OTnnc4y1VgAAAADC1yDuAgAAAACVguAbAAAAiAjBNwAAABARgm8AAAAgIgTfAAAAQEQaxV2AKLVt29Z26tQp7mIAAACgnhs7duxsa2279PkVFXx36tRJY8aMibsYAAAAqOeMMd9lmk+zEwAAACAiBN8AAABARAi+AQAAgIgQfAMAAAARIfgGAAAAIkLwDQAAAESE4BsAAACICME3AAAAEBGCbwAAACAiBN8AAABARAi+Q7b4v//Vt3vtpf1eeUUbjBqlhqNGqfXo0Wo3erRajBqlqlGjZDLcGo4apaajRqlVYtpi1Ci1GT1azUeNUuPErYlv/WajR+tfM2em7Hv+qlV64uefNX/VqqLKXur6+W4v03z/vNrKUUw581mn2O3eP2OGHpgxI6/1ij3G3y1bpgETJ+q7ZctK2k4h+8hHrnJkeq6QfZTyPn+3bJnunzFDg6ZPX/fe+J/zbzd9P0F/jsI690o5Jwr5vBVa5nzPiULLEEQ5g/rcFPMaC9lWMce2mOUKXTYfQWwvyPIH+friOlZBXpNK/dxmml/bNTXo15BvuctJo7gLUN99PmmS9nz3XX03f77mNW8uSZpvba3rrZW0MnFTYrokx3rLrdXZ33yjYzfaaN28l+bM0e8nT5Yknda+fcFlL3X9fLeXab5/nqSc5SimnPmsU+x2r5gyRZJU3bBhresVe4xvmj5dg3/6SZL0ULdugb9XmfaRj1zlyPRcIfso5X0+ql07DZ05U6usVWNjVN2woSSte27YrFnrtpu+n6A/R2Gde6WcE4V83vLZRm2f79r2m08Zin1fgnithZSl0P3Vdj3M99jmW75Sls1HENsLsvxBvr64jlWQ1yQp8zlZ6LXcP7+2a2rQr6GYz0TcCL5DtmOLFpKkrs2aaZ6kBZLWM0aNJC2zVqslrciwXgO5N6dK0nJJjSU1NUbLrV0XkJvEciskVRmjh7t2TdnG4W3apEwLVer6+W4v0/xc84IoZz7rFLvdJZ07y+S5XrHH+OoOHVKmQb9XmfaRj1zlyPRcIfso5X3u06qVdmzRQsvWrFF1w4Yp2+jTqpX2bNWqxvazTQtVzPaK2Wcp50Qhn7dCtlFbOWrbb9DHKJ99FqOU11jI8oUc22KWK3TZfASxvSDLH+Tri+tYhfHdXurn1v+4tmtqWK8hiO1Gxdg8amHri969e9sxY8ZEu9Mnn5ROO02aPFnq3DnafQMAACAWxpix1tre6fNp8w0AAABEhOAbAAAAiAjBd0Se+vlnDZo+XTdNnaozJkwoKIOElLsHb6FZKfLteezNG7doUcFZL/Ite65lC+21HEaP/nx6dReS3aTY8tT2HheTmSRXOdJ7qxe7rUzHJn1+mJktSlVq5pN8j2cYWROynZeFZINIv5+eMSbI1xF0FpJM790d06frjAkTNG7RonX3s2WHyfbYey+zvadhnqvFfHbS+a8VxWShyPQ+FXJMismiUegyhaz73bJlOn3CBA2aPj2v614p5Q/i9dV2LP3nSG3vR6nf7fmUv9RraG37CeK7Kg50uIzIrdOn65vVq7Um0cmySYMGeWeQkHL34C00K0W+PY+9eds1b663FyzIe/uFlD3XslJhmQjC6NGfT6/uoYkUj/lkNym2PLW9x8VkJslVjvTe6sVuK1Pml/T5UniZLUpVauaTfI9nGFkTsmXdKSQbhJSageSKKVNSMsbkmwUh3zIHmYUk03t39dSpWmGtpi1frg8XLtQKa9ddi2t7r9Pfy2zvaZjnajGfnXT+a8WerVoVnIUi0/tUyDHJVeawMgHlWvem6dP1xMyZamqMvl62rNbrXjFZQIJ8fbUdS/858unixTnfj1K/2/Mpf9DZo9L3E8R3VRwIviPyxw4dNHPzzbVszRpNWr68oAwSUu4evIVmpSg0u8gOzZvr3h9/LLjMubZdTNnC3k8hy6Vn0cg3u0mx5antPS4mM0mucqT3Vi92W5kyv+SaX1u5ou7BXmrmk3yPZxhZE7Jl3SkmG4R3f0nnzhkzxuS77drKnG2/xcj0ns3aYguNX7JEv99sM42cN0/jlyzJmh0m29R7L7O9p2Geq8V8dtL5rxWtGjWqsX4x50cxx6TY8yfoc+zqDh20Yu1a9WzeXEe3a1frda+YLCCFlL2Uz6d33ztHfrvBBjnfj1K/2/MpX6nX0Nr2E8R3VRzIdhI2sp0AAABUHLKdAAAAADEj+I5KBf3DAAAAgMwIvsNmTNwlAAAAQJkg+K4nok5/5E/blS1NUdApocJaL8gUhfmk7gorFVm+72Nt6eIK2Z5/m+npz/wpoII4BpnS3pWa4rGQfQa5vfT0WKXuJ5/jEcR1oJR1si1b7PUn3/2mp+EMOqVhoYK83hS7bPo65ZSuLdf7E8f7FaRiX0+h6wV93IrZXj7fhbUtX6pyPl/IdlJPRJ3+yJ+2K1vaxKBTQoW1XpApCvNJ4xhWKrJ838fa0sUVsj3/NqXU9Gf+FFDePG9bpb5PkrKm0gtSGCkAM6XHKnU/uVILpu872z7i+uwUe/3Jd7/paTiDTmlYqDBSoha6bPo65ZSuLdf7E1fK0aAUex4Xul7Q3znFbC/flMa5li9VOZ8vBN/1RNTpj/xpu7KlKQo6JVRY6wWZojCfNI5hpSLL932sLV1cIdvzbzM9/VmmFFClHIP0dbKl0gtSGCkApZrHptT95EotmL7vINMdBvHZKfb6k+9+09NwBp3SsFBBXm+KXTZ9nXJK11bINbSuKfY8LnS9oL9zitlePu9jbcuXqpzPF1INhu2f/5ROPVWaNEnq0iXafQMAACAWpBoEAAAAYkbwDQAAAESE4LueC6K3b209lOtS1omge32HrdTe4XH39s43+0s+GVhKzcwQ97EoVhBZSuLKJBDU9ce7xkTx3heS2SHM60m2a2s5nMfFZukoB6Wct2FlXynl2paezSfffQX1npTje1wX0OGynguit29tPZTrUtaJoHt9h63U3uFx9/bON/tLPhlYSs3MEPexKFYY2YqiKFdQ+/VfYz5dvDj0976QzA5hXk+yXVvL4TwuNktHOSjlvA0rc1Up17b0bD757quQfUS5vUpB8B2VmDq2BtHbt7YeynUp60TQvb7DVmrv8Lh7e+ebuSCfDCylZmaI+1gUK4xsRVGUK6j9+q8xv91gg9Df+0IyO4R5Pcl2bS2H87jYLB3loJTzNugsIunbLebalp7NJ999BfWelON7XBeQ7SRsTz0lnXKK9M030lZbRbtvAAAAxIJsJwAAAEDMCL4BAACAiBB8I0WmXtdh9PYvNINDtqwZYWZaCTtTTJDC6oVf6LaieL1hZp8Iogz5rFNKBpO4swvUlgmi2Iwkpe4/im3mk70njP0Ws71cWTCKea/Cft/LJRNVpu+VoN/3YssRhfTrVL5lCOKaFdT3WNzXyHzQ4RIpMvW6lhR4b/9CMzhky5oRZqaVsDPFBCmsXviZth3UsmGUJ67jXeg6UvbPVG3bjzu7QG2ZIIrNRlPq/qPYZj7Ze8LYbzHbe3fBgqxZMIp5r8J+38slE1Wm75Wg3/diyxGF9OtUvmUI4poV1PdY3NfIfBB8I0WuXtdB9vYvNINDtqwZYWZaCTtTTJDC6oVf6LaieL1hZp8IogyFrFNMBpO4swvUlgmi2Gw0pe4/im3mk70njP0Ws70+rVpJypwFo5j3Kuz3vVwyUWX6Xgn6fS+2HFFIf135liGIa1ZQ32NxXyPzQbaTsJHtBAAAoOKQ7QQAAACIGcE3AAAAEBGCb6QotJdwOfcqrq3ndBC9yaPqkR5Gdo2gyl5IJoA4sgbkEtb5Xsjrj1u2MuXKllHIdrI9F2R2iriyn5SzbJ/vQrLulIswPqelZBwqRl051vnK9/UEeR2pT+hwGZU60ra+0F7C5dyruLae00H0Jo+qR3oY2TWCKnshmQDiyBqQS1jneyGvP27ZynTT9OlZs2UUsp1szwWZnSKu7CflLNvnu5CsO+UijM9poRm3SlUfzim/fF9PkNeR+oTgO2zGxF2CghTaS7icexXX1nM6iN7kUfVIDyO7RlBlLyQTQBxZA3IJ63wv5PXHLVuZvCwZmbJlFLKdbM8FmZ0iiO2Evc2oZft8F5J1p1yE8TmtbZkwM9HUB/m+niCvI/UJ2U7C9vTT0sknSxMnSl27RrtvAAAAxIJsJwAAAEDMCL4BAACAiBB81xO1Za4oNUtDUGWsLQNEKT3xg+jF76333bJlOTOlFPL6itl/pmOU/v7G8Z6Gnd2lmNcURQaVfDIjlPr5K8de/nGXu5iMFKVmmonqfYgiI045nlNBKvX1BXl+x5XhyLv2DJo+XYOmT895bU4vS1hli/JYhP2dFBY6XNYTtWWuKDVLQ1BlrC0DhKSie+IH0Yvf28ZR7dpp2KxZ69YPovd8IftP30am9zeO9zTs7C7FvCYpcyaXKDMVBPH5K8de/nGXu5iMFKVmmonqfYgiI045nlNBKvX1BXl+x5XhyLv2rEr032tsTNZrc20ZwIIsU1THIqqMY0Ej+K4nastcUWqWhqDKmGnb+WTlKHb7hb4Ob/k+rVppz1atsmZKyXf/hcp1jNLf3zje07Czu5TymsLMoJJPZoRSP3/l2Ms/7nIXk5Gi1EwzUb0PUWTEKcdzKkilvr4gz++4Mhx5155la9ZIcgFovp+XsMoW5bEI+zspLGQ7CRvZTgAAACoO2U4AAACAmBF8AwAAABEh+I5KTM174urtXmgPcX92ke+WLdOAiRP13bJlEZU2P9kyocS5r/SsH/5e34W890H3Tg/iWOXadz699uM498vt3K3v2S7CUGpGiLgzxFSCMDNLFbOfcsk8les6mOtaHFYWkrjP8XLOhEKHy7DFPLx8XL3dC+0h7s8u8u6CBRr800+SpIe6dQu/sHnKlgklzn2lZ/3w9/qW8s/4EnTv9CCOVa5959NrP45z/6bp08vq3K3v2S7CUGpGiLgzxFSCMDNLFbOfcsk8les6mOtaHFYWkrjP8XLOhELwXc/F1du90B7i/uwifVq1kiRd3aFDuIUsULZMKHHuK/04Z+r1nU85g+6dHsSxyrXvfHrtx3Hue+dsuZy79T3bRRhKzQgRd4aYShBmZqli9lMumadyXQdzXYvDykIS9DaLKUO5ZkIh20nYhg6VTjpJ+vprqQxqwgAAABA+sp0AAAAAMSP4rkei6OxVSieKcuiAEbVCOg3WtnwUwuhcFra6dk5GNfRyMduLerjpQtYPs+NUbeVL79hcSkfmOD5jUQxxXsq5Vg6v0d8xMde5Fvc1r9RO8N565dY5vDZxH/eg0ea7Homis1cpnSjKoQNG1ArpNFjb8lEIo3NZ2OraORnV0MvFbC/q4aYLWT/MjlOFDGMvldaROY7PWBRDnJdyrkn5H9Mgy5Bp/aPatdPQmTMlZT7X4r7mldoJ3luv3DqH1ybu4x40gu96JIrOXqV0oiiHDhhRK6TTYG3LRyGMzmVhq2vnZFRDLxezvaiHmy5k/TA7TpUyjH2h243jMxbFEOeFbr/YYxpkGTKt36dVK+3YokXWcy3ua16pneC9abl1Dq9N3Mc9aHS4DBsdLgEAACoOHS4BAACAmBF8R6WC/mEAAABAZgTfYYtwhMvaetYH1cvePz/fHtNBDdmc73DlxQ73m8/rCWLo9KCUa3aKsLeXbR+lZMSIsjd9VMcjzOGz87neFPp+BD38dhDbivK98l9T4s5WU9u2Cs2Wkel8KPV7IZ/yFzKseq7vvGKvLYWuG9S5UKhyyDhTV/YZBDpc1iO19ayXCutRnk9WhnyHgg9qyOZ8hysvdrjffHqARznMfG3KNTtF2NvLto9SMmJE2Zs+quMR5vDZ+VxvCn0/gh5+O4htRfle+a8pUuEZQKI8foVmy8hnOPQwyl/IsOq5vvOKvbYUum5Q50Kh6kJWnXLZZxAIvuuRfHvWl9rL3j8/36HggxqyOd/hyosd7jefHuBBDJ0elHLNThH29rLto5SMGFH2po/qeNS2j1KGxc5nXqHvR9DDbwexrSjfq0zXlLiy1dS2rUKzZWT6fJb6vVDb/qTChlXP9Z1X7LWl0HWDOhcKVRey6pTLPoNAtpOwPfOMdOKJ0oQJUvfu0e4bAAAAsSDbCQAAABAzgm8AAAAgIgTf9UiuXr+l9ggOc9uFbiPILApx9ZSuqz20g1JXXn+Y5Qw7M0Up2w/jdZeSfaWcsu8UkyEpyP2Xw3ZKeZ/iOreC3m45ZkeKMkNO1NeXoM7RckHwXY94vX5fmjOnoOfi3nah2/B6j18+ZUre+8y2/SDKXoy49lsu6srrD7OcQW4707ZK2X4YrzufbUb1OQ3y2HhZP26aPj2S/ZfDdkp5n+I6t4LebpTXsHz3VUyZin0dUV9fgjpHywXZTuqRXL1+S+0RHOa2C91GMT3P88ncEqW62kM7KHXl9YdZzrAzU5Sy/TBedynZV8op+04xGZKC3H85bKeU9ymucyvo7ZZjdqRiylTs64j6+hLUOVouyHYSNrKdAAAAVByyncStgn7kAAAAIDOC77BFOLw8AAAAyhvBN8pe0Jkc4lRoVpewergHvZ1y6Ile7udEuZcvl3Iqe3pZorg+RP1ZLIVXhu+WLYu9LMXIN5tVFJmI/Mcw7EwucWWKyVel7jssBN8oe0FncohToVldwurhHvR2yqEnermfE+VevlzKqezpZYni+hD1Z7EUXhlumj499rIUI99sVlFkIvIfw7AzucSVKSZflbrvsJDtBGUv6EwOcSo0q0s+y5ZD5oRy6Ile7udEuZcvl3Iqe3pZorg+RP1ZLIW37z6tWmnPVq3K4j0rRL7ZrKLIRJTpGIZVprgyxeSrUvcdFrKdhO3ZZ6UTTpC++kraeuto9w0AAIBYkO0EAAAAiBnBdz3x7rx56vj+++rw3ntqPmqUen3yibb44AP9YdKkjB1X5q9apTumT9cZEyYUNCyyt262Dk9hd4wIo8NiKdspZfjhQo5bVMNuZ9pmXe7wWls5Mw0N7l+nkKHDazsX8ulElo9iO9RFPRx0KduM8nzPR1CdGMvhPSj2WlOfZLr25npvS7nOF1OuQs6zIK8thUgva1DXI++aO27RorLugFoq2nzXE/2/+UbTV65c9/jTJUskSYNmzFDzBg1U3bChTmvfft3zL82Zo6unTtUKa9WkQQM91K1b3vvyOj9I0mnt26c8lpTyXNDS9x3UssVup9jn0p+Xch+3bNsK6jXm2lemfYSx3zDUVk5vaHBJ6z4D/nXeXbCgxvPF7MvrRCapxmexUN5+jmrXTsNmzcq4v0LLF+a6xWwzyvM9H8Ue82zbKWb9KK5nQe6nnGW69uZ6b0u5zhdTrkLOsyCvLYVIL2tQ1yPvmvzN0qX6MhHHZNteXT5XCb7ricFdu+rkCRNkrdWcVavUvXlzzV29Wke3basu1dUZhwGetcUWGr9kSUHDInvr5pqm3w9SGB0WS9lOKcMPF3Lcohp2O9M263KH19rKmWlocP86fVq1qvF8MfvKtxNZPortUBf1cNClbDPK8z0fQXViLIf3oJjrUn2T6TXmem9Luc4XU65CzrMgry2FSC9rUNcj71p74Sab6PMlS8q2A2qp6HAZNjpcAgAAVBw6XMatgn7kAAAAIDOC77AxvDwAAAASCL7rmaiHFq7LvY0BAACiRofLeiaoXvmF7i/s/QAAANQHBN/1TNRDC9fl3sYAAABRI/iuZ9Zv3HhdDfRpzZpFuj8AAADkRptvAAAAICIE3wAAAEBECL4BAACAiBB8AwAAABEh+AYAAAAiQvAdFYaXBwAAqHgE32FjeHkAAAAkEHwDAAAAESH4BgAAACJSVsG3MabaGPOtMcYaY+7N8Hw3Y8xwY8w8Y8wSY8w7xpj94igrAAAAUKiyCr4l3SCpXaYnjDGdJb0vaXdJt0m6XFILSa8ZY/aPrIQAAABAkcom+DbG7CTpEknXZVnkZknrSzrIWnuztfYfkvaS9KOk+4yhZyMAAADKW1kE38aYhpIelvRfScMyPN9c0uGSRllrP/fmW2sXSxosqauknSMpLAAAAFCksgi+JV0qqbukC7M830NSU0kfZHjuw8SU4BsAAABlLfbg2xizhaTrJd1grZ2WZbFNEtMZGZ7z5m2aZfsDjDFjjDFjZs2aVVJZAQAAgFLEHnxLekDSt5IG5VimOjFdkeG55WnLpLDWPmSt7W2t7d2uXca+nNFghEsAAICK1yjOnRtjTpF0gKS9rbWrciy6NDFtmuG5qrRlygv9QAEAAJAQW/BtjGkqV9v9qqSfjTFdEk95zUdaJebNlsto4n/Oz5uXqUkKAAAAUDbibHbSTC6n9yGSJvluoxLPn5J43F/SeLkmJ7tn2M5uiemYEMsKAAAAlCzOZidLJB2bYX47Sf+QSzv4iKRx1trFxpj/SDrKGNPTWvuFJBljWsgF55MkfRxNsQEAAIDixBZ8J9p4/zt9vjGmU+LuFGut//mrJP1a0ghjzJ2SFko6W67ZySHW0qMRAAAA5S3WDpeFsNZONsbsIekWSVdKaiLpU0m/sda+HmvhAAAAgDyUXfCdyPWdMUWItXaCpL6RFggAAAAISDnk+QYAAAAqAsE3AAAAEBGCbwAAACAiBN9RIRkLAABAxSP4DhvDywMAACCB4BsAAACICME3AAAAEBGCbwAAACAiBN8AAABARAi+AQAAgIgQfAMAAAARIfgGAAAAIkLwDQAAAESE4BsAAACICMF3VBheHgAAoOIRfIeN4eUBAACQQPANAAAARITgGwAAAIgIwTcAAAAQEYJvAAAAICIE3wAAAEBECL4BAACAiBB8AwAAABEh+AYAAAAiQvANAAAARITgOyoMLw8AAFDxCL7DxvDyAAAASCD4BgAAACJC8A0AAABEhOAbAAAAiAjBNwAAABARgm8AAAAgIgTfAAAAQEQIvgEAAICIEHwDAAAAESH4jgojXAIAAFQ8gu+wMcIlAAAAEgi+AQAAgIgQfAMAAAARIfgGAAAAIkLwDQAAAESE4BsAAACICME3AAAAEBGCbwAAACAiBN8AAABARAi+AQAAgIgQfEeF4eUBAAAqHsF32BheHgAAAAkE3wAAAEBECL4BAACAiBB8AwAAABEh+AYAAAAiQvANAAAARITgGwAAAIgIwTcAAAAQEYJvAAAAICIE3wAAAEBECL6jwvDyAAAAFY/gO2wMLw8AAIAEgm8AAAAgIgTfAAAAQEQIvgEAAICIEHwDAAAAESH4BgAAACJC8A0AAABEhOAbAAAAiAjBNwAAABARgm8AAAAgIgTfUWF4eQAAgIpH8B02hpcHAABAAsE3AAAAEBGCbwAAACAiBN8AAABARAi+AQAAgIgQfAMAAAARIfgGAAAAIkLwDQAAAESE4BsAAACICMF3VBjhEgAAoOIRfIeNES4BAACQQPANAAAARITgGwAAAIgIwTcAAAAQEYJvAAAAICIE3wAAAEBECL4BAACAiBB8AwAAABEh+AYAAAAiQvANAAAARITgOyoMLw8AAFDxCL7DxvDyAAAASCD4BgAAACJC8A0AAABEhOAbAAAAiAjBNwAAABARgm8AAAAgIgTfAAAAQEQIvgEAAICIEHwDAAAAESH4BgAAACJC8B0VhpcHAACoeATfYWN4eQAAACQQfAMAAAARIfgGAAAAIkLwDQAAAESE4BsAAACICME3AAAAEBGCbwAAACAiBN8AAABARAi+AQAAgIgQfAMAAAARIfiOCsPLAwAAVDyC77AxvDwAAAASCL4BAACAiBB8AwAAABEh+AYAAAAiQvANAAAARITgGwAAAIgIwTcAAAAQkViDb2NMN2PMU8aYCcaYBcaYpcaYr40xg4wxG2dZfrgxZp4xZokx5h1jzH5xlB0AAAAoVKOY97+ZpI0lvSDpB0mrJW0vaYCkE4wxO1hrf5EkY0xnSe8nlrlN0gJJZ0t6zRjzW2vt6zGUHwAAAMhbrMG3tfYNSW+kzzfGvC3pOUlnyAXaknSzpPUl9bLWfp5Y7glJ/5N0nzGmu7VlPIxkGRcNAAAA0SjXNt/fJaatJckY01zS4ZJGeYG3JFlrF0saLKmrpJ0jLmN+GOESAAAACWURfBtjqowxbY0xmxljDpT0YOKpVxPTHpKaSvogw+ofJqblGXwDAAAACWURfEvqL2mWpO8lvSbXvOQUa+07iec3SUxnZFjXm7dppg0bYwYYY8YYY8bMmjUruBIDAAAABYq7w6VnuKSvJbWQtKNcE5O2vuerE9MVGdZdnrZMCmvtQ5IekqTevXvT8BoAAACxKYvg21r7g1y2E0kabox5XtInxphqa+3NkpYmnmuaYfWqxHRphucAAACAslEuzU5SWGvHSfpM0vmJWT8mppmalnjzMjVJAQAAAMpGWQbfCc0kbZC4P16uycnuGZbbLTEdE0WhAAAAgGLFPcJl+yzz95W0nRKZTBIpBf8jaR9jTE/fci3kOmtOkvRx6AUGAAAAShB3m+/7E8PIvymX27tKUi9JJ0haJOky37JXSfq1pBHGmDslLZQb4XJTSYeU9QA7AAAAgOIPvodKOk3SqZLaSbJyQfiDkm631k73FrTWTjbG7CHpFklXSmoi6VNJv2FoeQAAANQFcQ8v/5zcMPL5Lj9BUt/wShQiKuYBAAAqXjl3uKwfGF4eAAAACQTfAAAAQEQIvgEAAICIEHwDAAAAESH4BgAAACJC8A0AAABEhOAbAAAAiAjBNwAAABARgm8AAAAgIgTfAAAAQEQIvqPC8PIAAAAVj+A7bAwvDwAAgASCbwAAACAiBN8AAABARAi+AQAAgIgQfAMAAAARIfgGAAAAIkLwDQAAAESE4BsAAACICME3AAAAEBGCbwAAACAiBN9RYXh5AACAikfwHTaGlwcAAEACwTcAAAAQEYJvAAAAICIE3wAAAEBECL4BAACAiBB8AwAAABEh+AYAAAAiQvANAAAARITgGwAAAIgIwXdUGOESAACg4hF8h40RLgEAAJBA8A0AAABEhOAbAAAAiAjBNwAAABARgm8AAAAgIgTfAAAAQEQIvgEAAICIEHwDAAAAESH4BgAAACJC8A0AAABEhOA7KgwvDwAAUPEIvsPG8PIAAABIIPgGAAAAIkLwDQAAAESE4BsAAACICME3AAAAEJFGcRcAAADUfytWrNDcuXO1aNEirVmzJu7iAAVr2LChWrZsqQ022EBNmzYtejsE3wAAIFQrVqzQ9OnT1bp1a3Xq1EmNGzeWIRsY6hBrrVatWqWFCxdq+vTp6tChQ9EBOM1OAABAqObOnavWrVurbdu2atKkCYE36hxjjJo0aaK2bduqdevWmjt3btHbIvgGAAChWrRokdZbb724iwEEYr311tOiRYuKXr/oZifGmGpJnSS1kVTjJ6y19u2iSwUAAOqNNWvWqHHjxnEXAwhE48aNS+q3UHDwnQi6B0k6M8v6RpKV1LDoUtVHDC8PAKhgNDVBfVHquVxMzffdks6S9KqkNyXNKakE9R0XGwAAACQUE3wfKWmotfbkoAsDAAAA1GfFdLiskjQq4HIAAAAgi4EDB8oYo2nTpsWy/zPOOIOmQwEpJvgeI2mroAsCAABQ140aNUrGmKy3Ro0YYqXSFXMGXCnpP8aY56y1Y4IuEAAAQF134okn6uCDD64xv0GD4rI8X3PNNbryyitLGlkR5aGY4HuApB8kfWiM+UDSt5LS861Ya+1ZpRYOAACgLtppp510yimnBLa9Ro0aUWteTxTz8+sMST0T6+4h6dTEvPQbAAAAMpg2bZqMMRo4cKCGDh2qHj16qKqqSh06dNDAgQO1evXqlOUztfmeO3euLr30UnXu3FlVVVVq06aNevXqpdtvvz1l3dWrV+vWW2/VNttss265I488UuPHj69RruXLl+vyyy/XJptsombNmmmXXXbRiBEjsr6OSZMm6dRTT9XGG2+sJk2aqFOnTrr88su1ZMmSlOW+//579evXTx07dlTTpk214YYb6le/+pUef/zxIo5e3VbwTyhrLaNiAgAA5LB06VLNnj27xvwmTZqkjPb50ksv6dtvv9UFF1yg9u3b66WXXtL111+v7777TkOGDMm5j2OPPVZvv/22zj33XPXo0UPLli3ThAkTNGrUKF1++eXrljv55JP13HPP6YADDtB5552nn3/+Wffdd5923313vfPOO9pxxx3XLXviiSdq+PDhOuyww3TQQQdpypQpOuqoo7TFFlvU2P/YsWO13377af3119c555yjTTfdVF988YX+/ve/67333tPo0aPVuHFjrV69WgcccIBmzJih888/X127dtWCBQs0btw4vfPOOzr99NOLOcR1l7U275ukFpIelXRsIeuVy61Xr142cm+8Ya1k7ejR0e8bAIAy8NVXX8VdhMi89dZbVm6wwYy3Qw45xFpr7dSpU60k26BBAzt27Nh1669du9YeccQRVpL94IMP1s2/7rrrrCQ7depUa6218+fPt5Lseeedl7M8I0aMsJLscccdZ9euXbtu/ueff24bNmxo99xzz3XzXnvtNSvJnn766SnbeOGFF9aV369Hjx62W7duduHChSnzhw0bZiXZIUOGWGut/eKLL6wke+utt+Y+eHVIPue0pDE2QzxaUC22tXaxpBMkrVfbsgAAAFGYv2qVnvj5Z81ftSruoqwzYMAAjRw5ssbtpptuSlnugAMO0E477bTusTFGV1xxhSTphRdeyLr9Zs2aqWnTpvroo49yph/0tnH11VenpArs2bOnDjvsML377ruaNWuWJGn48OGSlFJrLklHHHGEunXrljJv/PjxGjdunE466SStWLFCs2fPXnfbc8891bx583XNVVq1aiVJeuutt/TLL79kLWulKKYJyVeSOgVcjvqP4eUBAAjFS3Pm6PeTJ+ulOeUz6PZWW22l/fffv8atZ8+eKcttvfXWNdbdZpttJEnffvtt1u03adJEd911l7788kttscUW2nbbbXXRRRfpjTfeSFlu6tSpatCgQcb9bLvttuuW8fbXoEEDde3atcay6etPmDBBknTdddepXbt2KbcNN9xQS5Ys0cyZMyVJHTt21NVXX60RI0Zo4403Vq9evXTFFVfok08+yfr66rNius3eJukfxpgnrbXfBF2geoeE9AAAhOrwNm1SppXi3HPPVd++ffXKK69o9OjR+ve//617771Xxx9/vJ555plQ920TlYqXXXaZfvOb32RcpnXr1uvu/+Uvf1G/fv30yiuv6J133tHgwYN1++2364orrtCtt94aalnLTTHBd3dJ30sab4x5WdIkSUvTlrHW2htLLRwAAEBt1m/cWKe1bx93MYri1SD7ffXVV5KkLbfcstb1N954Y/Xv31/9+/fXmjVrdOqpp2ro0KG67LLLtPPOO2vLLbfU2rVrNWHCBPXo0SPjfrzOlN6y33zzzbpa8Wzl3GorN95iw4YNtf/+++f1WrfccktddNFFuuiii7R8+XIddNBBuu2223TZZZdpww03zGsb9UExzU4GyqUabCzpSElXJOal3wAAAJDDyJEj9emnn657bK3VbbfdJsm1tc5m6dKlWro0te6zYcOG6wLsuXPnpmzj5ptvXldbLUlffvmlXnrpJe25555q166dJKlv376SVCNV4fDhwzVx4sSUeTvuuKO22247PfDAAxmbx6xevXpdGRYsWKBVae3xq6qq1jVlmTdvXtbXWR8VU/NdM9cMAAAA1vn000/1z3/+M+Nz/qC6Z8+e2m+//XTBBRdo44031osvvqjXX39dp556qnbfffes2//mm2/Up08fHXnkkdpuu+3UunVrTZgwQffff7+22GIL7bXXXpJch87jjjtOzzzzjObNm6dDDz10XarBqqoq/f3vf1+3zYMOOkiHHXaYHn/8cc2dO1e/+c1vNGXKFD344IPabrvt9OWXX65b1hijJ598Uvvtt5969Oihfv36adttt9XSpUs1efJkDRs2TDfffLPOOOMMvfXWWxowYICOPvpodevWTS1atNDYsWM1ePBg7brrrjU6c9Z7mVKg1NdbLKkG33zTpRocNSr6fQMAUAZINZh6mzRp0rpUg9ddd519+umn7fbbb2+bNGliN9tsM3vttdfalStXpmw3PdXg7Nmz7SWXXGJ79uxpW7VqZauqqmznzp3txRdfbH/88ceUdVetWmVvueUW2717d9ukSRPbunVr27dvXztu3Lga5V+6dKn9/e9/bzfaaCNbVVVld955Z/vaa6/Z008/vUaqQWutnTZtmj3nnHNsx44dbePGje0GG2xgd9ppJ3vllVfa6dOnW2ut/fbbb+0555xju3fvblu2bGmrq6tt9+7d7bXXXmvnz58f0JGPVimpBo2toCwcvXv3tmPGjIl2p2+9Je23nzRqlNSnT7T7BgCgDEyYMCFjto1KNm3aNG2xxRa67rrrNHDgwLiLgwLlc04bY8Zaa3unzy+42Ykx5tE8FrPW2rMK3TYAAABQnxXT5vuMPJaxkgi+AQAAAJ+Cs51Yaxuk3+Qyn3ST9LCkDyW1zrkRAAAAoAIVU/Ndg7V2jVy+73OMMf+RdKuk84LYdr0RZtv6F1+UXn1VevDB8PYBAAAC06lTJ1VSvzskFZPnuzb/lXR0CNutmwod4fLzz906J54orV2b3zpHHCE99JC0ZEmhpQMAAECEwgi+N5DUIoTtVoYbbnDTZ56RJk8ubN3ly4MvDwAAAAITWPBtjFnfGHOMpEsljQ1quxWt0KTzaSNdAQAAoLwUHHwbY9YaY9ak3yTNkfScpLWSfh90QStGevuvBQvyX/eee2h6AgAAUMaK6XD5hFwqQT8raa6kbyQNtdYuKrVgFWv48NTHvXrl3/zk9ttdsE7HSwAAgLJUcPBtrT0jhHIgmylTClt+6tRwygEAAICSFdPs5M/GmO1yPL+tMebPpRULRVu5Mu4SAAAAIItiOlwOlNQjx/PbSbquqNJAatu2tPV/+imYcgAAACBwYaQarJK0OoTtVoYuXaQDDkgG0fkE402bJu8vork9AAB1RadOnbTPPvukzNtnn33UqVOnvNYfNWqUjDF67LHHcs5D+cgr+DbGrGeM6WCM6ZCY1cZ7nHbbQdLJkr4Pq8D13rJlUnW11L69dOyxLnf3nDnZl7dWWrVK6t/fPS50UB8AABCob7/9VgMGDFD37t1VXV2t1q1ba+utt9bpp5+ut956K+7iReaMM86QMUbGGI0ZMybjMnfeeee6ZSrlx0K+HS4vleS147aS7krcMjGSriipVPVRPkPIrlkj/fyztPXW7vHzz7tRLjfc0D2XyerVbpmOHaXOnV0HzcmTXQ06AACI1JgxY9SnTx81btxYp512mrbddlstW7ZMkyZN0ogRI9SyZUvtu+++ObcxYsSIkoae33vvvbVs2TI1bty46G0EqaqqSkOGDFHv3r1rPDdkyBBVVVVpeQUNFJhv8D0qMTVyQfgLksalLWMlLZb0obX2/UBKVx8UUhPdqFHq1BtePtcw897JWlWVbHJy+eXSCy8UVk4AAFCy66+/XkuXLtXnn3+unj171nj+559/rnUbTZo0KakMDRo0UFVVVUnbCNKRRx6poUOHatCgQWrqayr7ySefaPz48TrppJP09NNPx1jCaOXV7MRaO9pae721dqCkxyXdlnjsv91grR1E4B2SxYszz/cH382aufvpucIBAEAkJk2apDZt2mQMvCWpffv2tW4jW5vvF198UTvuuKOqqqq0+eab69prr9WqVatqLFdbO/AhQ4Zo2223VdOmTdWxY0fddtttGctx//33q1u3bmratKm22mor3XvvvXrsscdkjNGoUaNqfR2eM888U/PmzdPwtPhkyJAhateunQ499NCM61lrdf/996tXr16qrq5WixYttO+++2ZsuvOPf/xDBx54oDbddFM1adJEG2+8sU455RRNmzatxrLGGJ1xxhn64IMP1KdPHzVv3lxt2rRR//79tThbvBWgYvJ8nxlGQeDzf//npocdJv3nP+5+y5aZm66sWOGmVVXSffdJ3gm8apVUJn83AQBQKTp37qyJEydq2LBhOuqoowLb7gsvvKCjjz5anTp10p///Gc1atRIQ4YM0SuvvFLQdh544AHNnDlTZ511ltZff33985//1B//+EdtttlmOumkk9Ytd+utt+rKK6/UTjvtpJtvvllLly7V7bffrnbt2hVc9h133FE77LCDHn30UR1//PGSpOXLl2vo0KE688wzszaPOfXUUzV06FAdc8wxOvPMM7VixQo99dRTOuCAAzRs2DAdfvjh65b929/+pt12202/+93vtMEGG+jLL7/U4MGD9eabb2r8+PFq06ZNyrY///xzHXrooTrzzDN10kknadSoUXrkkUfUoEEDPfTQQwW/xkIUM8KljDEt5dqBHyhpI0mnWWs/MMa0lXS+pOestV8HV8wK4P/l6nWw/MtfksG3JC1d6jpjStKwYdJvfiP997/ucVWVdMghyWWXLSP4BgCUt0sukT7/PO5SpNphB+muu4pe/ZprrtHIkSN19NFHa6utttKee+6pnXfeWfvss4+29vp0FWjNmjW6+OKLtcEGG+jjjz9W20QmtHPOOUc9euTK/lzT9OnTNWHCBLVq1UqS1K9fP3Xs2FH33HPPuuB77ty5GjhwoLbffnu9995765qw9O/fX926dSvqNfTr10+XXHKJfvjhB2222WYaNmyY5s+fr379+unrr2uGjC+88IKeeuopPfjggxowYMC6+RdffLF22203XXzxxTrssMNkEs17x48fr+bNm6ds4/DDD9f++++vRx55RFdckdodcdy4cfrggw+06667SnLHcuHChRoyZIgGDRqkFi1aFPU681HMIDvtJI2RdK2kNpK2lNRMkqy1syWdLmlA1g0gs0wdDbbaKvXxb3/rpmPHSkcfLV10kXT22W5ew4ZuevLJbrpkSTjlBAAAWe2+++4aO3asTj/9dC1YsEBDhgzR+eefr2222UZ77723vv3224K3OXbsWH3//fc688wz1wXektSqVSude+65BW3rzDPPXBd4S1J1dbV22203TZo0ad28kSNHavny5TrvvPNS2o63b99eJ3txRoFOPvlkNW7cWI8//rgk1+Rk55131nbbZR638Z///KdatmypI444QrNnz153mz9/vg477DBNmzYtpcxe4L127VotWLBAs2fPVs+ePdWqVSt99NFHNba/++67rwu8Pfvtt59Wr16dsalKkIqp+f6LpPaSdpU0XdIvac+/KOnXJZar8nz3XfJ+r15u2qyZ1LOn9MUX7vHbb7up1x7Jd9LJaxt24IHSU0+5WnIAAMpZCTXM5Wz77bdf1976u+++0+jRozV48GC988476tu3r8aOHVtQp0ovYO/evXuN57bZZpuCyrblllvWmNemTRvN8aU1njp1qiRlrOUutuZ7gw020OGHH67HHntMp5xyit58803de++9WZefMGGCFi1apI022ijrMjNnzlTXrl0lSW+++aZuuOEGffTRRzUyp8ybN6/GutmOg6SUYxGGYoLvQyX9w1r7qTGmTYbnv5V0RkmlqkTbb5+8v/vuyfuffSY1SPuDom9fN/UPJd+ypZt6zVIIvgEAiF3Hjh112mmn6dRTT9Vee+2l9957Tx9//LH23HPPWMrT0PunPAb9+vXTb3/7W5199tlq0qSJTjzxxKzLWmvVrl27nFlQvFrzTz75RAceeKC6dOmiW265RVtssYWaNWsmY4xOOOEErc2QNS7XcSglzWM+igm+20qanOP5tXKjXCJfzz2X+viOO5L3jZEmTpT8vzQXLHDTjh2lX36Rpk6VvL9tvPZONDsBAKBsGGO066676r333tOMGTMKWterpc3UNvqrr74KpHx+XqaViRMnar/99kt5buLEiUVv98ADD9Rmm22mkSNH6qSTTtL666+fddmtttpK33zzjXbbbbda218//fTTWrNmjf7v//5PW2yxxbr5S5YsyVjrHbdihpf/WVLnHM/vKNccBfkaMSJ5/5FHpPS/ohJ/qUiS/KM/TZ/uasX97a+8mu+5cwMvJgAAyG3kyJFavXp1jfnLli3TiMT3faFNRXr16qXNNttMQ4YM0ezZs9fNX7hwoR544IHSCpzBAQccoKZNm+r+++9PacLx888/66mnnip6uw0aNNB9992n6667Tn/84x9zLnvaaadp7dq1uuqqqzI+P3PmzHX3vVrs9Brrv/71rxlrveNWTM33q5LOMsbcI2ml/wljzK6STlP20S+RiX8gnm23zb3smb5MjzNnuo6aXn5vSfJS6XzyiXTwwcGVEQAA1OrSSy/VnDlzdPjhh2v77bdXdXW1vv/+ez399NP65ptvdNppp2l7f1PTPDRs2FB33nmnjjvuOO2yyy46++yz1ahRIz366KNq06aNpk8Pts6zTZs2uu666/SnP/1Je+yxh0455RQtXbpUDz30kLp27aoxY8asyzJSqMMPPzwlRWA2XnrBe++9V59++qkOPfRQtW3bVj/88IM++OADTZ48eV1b+COPPFJ33nmnDj74YA0YMEBNmjTRyJEjNW7cuJQOquWimOD7ekmHS/pM0ktyI1uebow5W9JRkn6SdGtgJawvcrUf8p/A+aYHPPRQF2CvWJGs7ZaSQ9MDAIDIDRo0SC+++KLeffddPf/885o/f75atWqlHj166I9//KPOOOOMorZ7zDHH6N///rduuOEGDRw4UBtuuKHOOOMM7b333jrwwAODfRGSrrrqKq233nq6++67deWVV6pDhw66/PLLZa3VmDFj1Mxf8ReSRx99VPvuu68eeugh3XzzzVq5cqXat2+/Lve4Z4899tDzzz+vG2+8Uddee62aNWum/fffX6NHj9bee+8dejkLZYppVG6M2VzSvZIOUbLpipWrFT/PWvtDYCUMUO/eve2YMWOi3enbb0t9+khvvCGltZta55xzJC+h+9q1mYeknzVL2nDD5OOLLpLuucfd328/t32PMVL//tLDDwfzGgAAKMGECROKznGN8nLRRRfp3nvv1U8//ZTXaJ31VT7ntDFmrLW2d/r8Ytp8y1r7vbW2r6QN5FIO7iapnbX2MEkdjTFv5NwAUvnbI2X7Gyd9RKn+/ZP333yz5vKDB5deLgAAUJHS0/VJ0k8//aQnnnhC2223XUUH3qUqqNlJIrVgZ0lzrbWTrbULJX2SeG43Y8wNcjm+y691ezlbubL2ZfzOPFMqsLMGAABAvkaNGqXLL79cRx11lDbbbDNNmzZNDz/8sBYvXqxbbrkl7uLVaXkF38aYhpLuk9RfkknM+0jSEZKWS3pA0vFyQffTkm4Koaz1V749cceNk3r0cHm+G/neumwfAmuz16QDAABk0aVLF3Xu3FkPP/yw5syZo6qqKvXu3VtXXXWV9t9//7iLV6flW/N9kdyQ8T9I+lBSF7mmJvdJ2kzSLpKelHSjtXZKCOWs37zc3D/+mHu57bd3HSzTUxGmp+vp00caPdrVqDdtGlw5AQBARejSpYuGDx8edzHqpXyD71MljZe0u7V2qSQZY+6TdJ6kOZL2tNZ+EE4RK8CDD7rpxhvXvmw+w9EecYQLvpcuJfgGAAAoI/l2uOwq6Qkv8E64PzG9lcA7Ju+/L334Yc35DDEPAABQlvKt+W4uN7Kln/d4fHDFQUF23z3zfIaYBwCUGWtt0QOzAOWkmDTdfoWkGkzfk/d4VUklQPC8mm+CbwBAGWjSpImWLVsWdzGAQCxbtkxNS2jWW0iqwYONMf6kjtVyAfixxpgd0pa11to7iy5VfZTrV1Lv3jXzeJdiq63cdOxYaccdg9suAABF8IYFb9u2rVq2bKlGjRpRC446xVqr1atXa9GiRZo9e7Y22mijordVSPB9UuKW7pwM86wkgm8pv1R/K1fm15EyX127uukvvwS3TQAAitSqVSs1bdpUs2bN0pw5c7R69eq4iwQUrFGjRqqqqlKHDh1UVVVV/HbyXG7foveA2o0bJy1cGNz2mjRxecAXLw5umwAAlKCqqkqbb7553MUAYpdX8G2tHR12QSrWmjVuOm1asNtdvVq6+Wbpr38NdrsAAAAoWiEdLhGGxx8Pd/v5jp4JAACA0BF8R+2nn1w7cC8/9+9+F+7+Fi0Kd/sAAADIG8F31P71LzcdNMhNvU4nF1wQzv4mTw5nuwAAACgYwXfULr7YTYcNc1Mv+G7ZMtj9jBzppgsWBLtdAAAAFI3gOy5eR0tvGnTb7LZt3fR//wt2uwAAACgawXeUcuU1DTr43mQTN13FAKQAAADlguA7SjfdlP25XCNgFqN1azdlOF8AAICyQfAdFWulSZOyPx90zXfjxlLDhgTfAAAAZYTgO2z+4eX99w84IHW5bbYJft/NmklLlwa/XQAAABQl1uDbGNPVGHODMeZDY8wsY8wiY8znxpirjTHNMyzfzRgz3BgzzxizxBjzjjFmvzjKXhR/8P3116nPnXVW8PurrqbmGwAAoIzEXfPdT9KlkqZIukHS5ZImSvqLpPeNMc28BY0xnSW9L2l3Sbcllm0h6TVjzP4Rl7s4/uD7+++zPxcUar4BAADKSqOY9/9vSTdba/3JqB8wxkySdLWksyTdm5h/s6T1JfWy1n4uScaYJyT9T9J9xpju1gbdazFg6QF22EO/U/MNAABQVmKt+bbWjkkLvD3PJqbbSVKiCcrhkkZ5gXdi/cWSBkvqKmnncEsbgPTge9UqqVs36bjjwtkfNd8AAABlJe5mJ9lslpjOTEx7SGoq6YMMy36YmJZ/8N0g7XCvWuVujRuHs79mzaj5BgAAKCNlF3wbYxpKulbSaklPJ2YnRozRjAyrePM2DblopZs/P/Xx6tXhBt/V1dL48eFsGwAAAAUru+Bb0l1ynSr/bK2dmJhXnZiuyLD88rRlUhhjBhhjxhhjxsyaNSvQghZs2LDUx6tWSStXSk2ahLO/+fOluF8zAAAA1imr4NsYc6OkCyU9ZK292feU13C5aYbVqtKWSWGtfcha29ta27tdu3bBFTYIYTc78XKJl3k/VAAAgEpRNsG3MWagpGskDZF0btrTPyammZqWePMyNUkpb7Nnhxt8N0+kSl+1KpztAwAAoCBxpxqUtC7wvk7S45L6Z0gZOF6uycnuGVbfLTEdE1oBg5Cp9rlnTzcNK/iuSvwpsHx5eE1bAAAAkLfYa76NMX+WC7yflNTPWlsj+XUipeB/JO1jjOnpW7eFpP6SJkn6OJoSFyifwXPCCoz9wTcAAABiF2vNtzHmAknXS5ou6XVJJ5nUYHWmtXZk4v5Vkn4taYQx5k5JCyWdLdfs5JCyH2Anl7Bqvr3OloMHS3/6Uzj7AAAAQN7ibnbi5ebuINfkJN1oSSMlyVo72Rizh6RbJF0pqYmkTyX9xlr7egRlDU9Ywbc3hP299xJ8AwAAlIFYg29r7RmSzihg+QmS+oZVntiEFXy3bOmmP/0UzvYBAABQkNjbfEPhBd8DB4azXQAAABSF4DsOG2+c+jisDpetWiXv1+Em8QAAAPUFwXccHk9r3v755+Hvc/bs8PcBAACAnAi+o+Kvee7TJ/W5hQvD3//554e/DwAAAORE8B0Vf/Cd3sY7zCYhBx/splOnhrcPAAAA5IXgOyprfWMHpQ+8s2ZNePvt1ctNW7QIbx8AAADIC8F3VNbWGLgzaf31w9vvpZe66QEHhLcPAAAA5IXgO2xeLXeu4Ltbt/D27wX2K1eGtw8AAADkheA7KtnadfftK/3+9+Ht1xipulpaujS8fQAAACAvBN9RyVbzfffdUsOG4e67WTNpyZJw9wEAAIBaEXxHJVvw3ahR+Ptu0kR6+eXw9wMAAICcIoj8ICl7s5Mogu/11gtvCHsAAADkjZrvqMRZ873zzjQ7AQAAKAME31GJs+a7RQtp8eLw9wMAAICcCL6jEmfNN8E3AABAWSD4jooXfA8YkDo/iuC7aVNp2TLpl1/C3xcAAACyIviOihd89+iROj+KjpAdOrjp9Onh7wsAAABZEXyHLX2ES+9xv35u2iCCt2CrrdyUTpcAAACxIviOitfh0gu2H3oounbYzZu7Ke2+AQAAYkXwHRWv5tsLvhs2TAbFYWvRwk0JvgEAAGJF8B2VCRPcdNy46PfdsqWbfv559PsGAADAOgTfURk2zE2ffTb6fW+6qZsyyiUAAECsCL6jsmJFfPtu0MA1caHZCQAAQKwIvqPyww9u6mU7idqSJdKdd0offRTP/gEAAEDwXXFGjIi7BAAAABWL4DtqcdV8e6ZOjXf/AAAAFYzgO2pxDfE+ZIibep0vAQAAEDmC70px+ukut7iXbxwAAACRI/gOW9zNTDzGkPEEAAAgZgTfUbvppvj23by5y3oCAACAWBB8Ry3OgW6aN5c+/DC+/QMAAFS4RnEXABFaskSaPFmytnyawwAAAFQQar6jFmfQO2CAmy5fHl8ZAAAAKhjBd9TiDL7btnVT2n0DAADEguA7au3axbfv6mo3Xbo0vjIAAABUMILvqB16aHz7bt7cTan5BgAAiAXBd9QaxHjIqfkGAACIFcF3JSH4BgAAiBXBd9SsjW/fXrOTt96KrwwAAAAVjOA7bOnZTeIMvnv0cNP58+MrAwAAQAUj+K4k1dVS+/Z0uAQAAIgJwXfU4hxeXnJNTxYvjrcMAAAAFYrgO2otW8a7/xYtqPkGAACICcF3pWnRQvrww7hLAQAAUJEIvivNsmXJrCcAAACIFMF3penVizzfAAAAMSH4rjQtWtDhEgAAICYE35XG63AZZ75xAACACkXwXWlatHCB988/x10SAACAikPwXWnWX99Np0yJtRgAAACViOA7bOnDy8fNG2Kedt8AAACRI/gOmz/47t49vnJ4vDSDBN8AAACRI/iOUjnUgrdo4aYE3wAAAJEj+A5bOQTcfl7wPWJEvOUAAACoQATflaZtWzdduDDecgAAAFQggu+wlVvNtzHSvvtK//tf3CUBAACoOATfUSqXgW3WrpWmTWOYeQAAgIgRfIfNX/O9dm185fA7+GA3XbAg3nIAAABUGILvSrTxxm66ZEm85QAAAKgwBN9h89d8l0uzE9INAgAAxILgO2zl2OzEG2hn/vxYiwEAAFBpCL6jVC7Bd+vWbvr++/GWAwAAoMIQfIetHJudbL993CUAAACoSATfUSqX4LtpU6lhQzpcAgAARIzgO2zlWPNtjLRmjfTXv0qffRZ3aQAAACoGwXeUyqXNt9+f/xx3CQAAACoGwXfYyrHm22/mzLhLAAAAUDEIvqNUTjXfjRu7KUPMAwAARIbgO2zlmOdbktq3d9NmzeItBwAAQAUh+I5SOTU7Wb3aTf0/DgAAABAqgu+wlWubby/4BgAAQGQIvsNWrs1ODjrITan5BgAAiAzBd5TKKfgePFjackupAacAAABAVIi8wlauNd9Nm0o9epDtBAAAIEIE31EqtwFtqqulcePK60cBAABAPUbwHTZ/zffOO8dXjkyaNHHTxYvjLQcAAECFIPiuZLvt5qYE3wAAAJEg+A5buaYalKQWLdyU4BsAACASBN9RKrfgu3lzN12yJN5yAAAAVAiC77D5a7432yy+cmTi1Xx/9lm85QAAAKgQBN9h8wffnTvHV45MunRx00WL4i0HAABAhSD4rmQbbeSmy5bFWw4AAIAKQfAdtnIevr2qyk0ZaAcAACASBN+VzBg30A7BNwAAQCQIvsNWzjXfktSsGc1OAAAAIkLwXemo+QYAAIgMwXfYyr3mm+AbAAAgMgTflY5mJwAAAJEh+A4bNd8AAABIIPgOW10Ivt94Q7I27pIAAADUewTfla5pUzddvjzecgAAAFQAgu+wlXvN90EHuSlNTwAAAEJH8F3pqqvdlOAbAAAgdATfYSv3mu9mzdyUjCcAAAChI/iudNR8AwAARIbgO2zlXvNN8A0AABAZgu9K5zU7mTcv3nIAAABUAILvsNWVmu9DD423HAAAABWA4DtsdSX4BgAAQOgIviudN8gOAAAAQkfwHbZyr/nu0CF5f8WK+MoBAABQAQi+K12jRtKdd7r7ZDwBAAAIFcF32Mq95lsi3SAAAEBECL4hNW/upkuWxFsOAACAeo7gO2zUfAMAACCB4BvUfAMAAESE4Dtsdanme+HCeMsBAABQzxF8h60uBN9t2rjp8OGxFgMAAKC+I/iG1L173CUAAACoCATfYasLNd/GSF27Sh99FHdJAAAA6jWCbzhLlkhNmsRdCgAAgHqN4DtsdaHmW5J23ZVUgwAAACEj+IZTXU3wDQAAEDKC77DVlZpvgm8AAIDQEXzDWb5cmjlTWrs27pIAAADUWwTfYasrNd9PPOGmI0bEWw4AAIB6jOA7bHUl+PbMmxd3CQAAAOotgm84m23mptbGWw4AAIB6jOA7bHWl5nvoUDdt2TLecgAAANRjBN9w2rZ102++ibccAAAA9Viswbcx5ipjzL+MMd8aY6wxZloty+9qjHndGLPIGLPQGPNfY8wO0ZS2SHWl5nujjdx05cp4ywEAAFCPNYp5/3+VNFfSp5LWz7WgMWY3SaMkzZD058TsCyW9Y4z5lbV2fHjFrACtWrnp8uXxlgMAAKAeizv47myt/VaSjDFfSmqRY9m/S1opaW9r7YzEOs9JmiDpDkkHhlzW4tSVmu8GDaSmTaVly+IuCQAAQL0Va7MTL/CujTGmi6SdJf3LC7wT68+Q9C9J+xtj2odTygrSrBnBNwAAQIjqSofLnRPTDzI896EkI6lXdMUpQF2p+ZbcEPME3wAAAKGpK8H3JonpjAzPefM2zbSiMWaAMWaMMWbMrFmzQilcTnUp+KbmGwAAIFR1JfiuTkxXZHhuedoyKay1D1lre1tre7dr1y6UwtUbBN8AAAChqivB99LEtGmG56rSlikv1HwDAAAgoa4E3z8mppmalnjzMjVJQSEIvgEAAEJVV4LvTxLT3TM8t5skK2lsdMUpADXfAAAASKgTwbe1drKkMZKONcZ4nS+VuH+spDettT/HVb56g+AbAAAgVLEOsmOMOVVSx8TDdpKaGGOuSTz+zlr7pG/xiyW9JTei5T2JeRfJ/YC4LIryFoWabwAAACTEPcLlWZL6pM27MTEdLWld8G2tfd8Ys4+kvyRuVtL7ko611n4RekkrAcE3AABAqGINvq21+xS4/AeSfh1OaUJCzTcAAAAS6kSb7zqtLgXf1dXS3LnS2rVxlwQAAKBeIvhGTVOmxF0CAACAeongO2x1qeZ7jz3cdPHieMsBAABQTxF8I6m62k2XludgoQAAAHUdwXfY6lLNN8E3AABAqAi+kdS8uZsuWRJvOQAAAOopgu+wUfMNAACABIJvJBF8AwAAhIrgO2x1qea7ZUs3nT073nIAAADUUwTfYatLwXerVlKTJm6gHQAAAASO4Bupqqul5cvjLgUAAEC9RPAdtrpU8y1JVVUE3wAAACEh+EYqgm8AAIDQEHyHrS7WfK9YEXcpAAAA6iWCb6Si5hsAACA0BN9ItXKl9O67cZcCAACgXiL4RqqvvnKpBj/4IO6SAAAA1DsE38jsxx/jLgEAAEC9Q/CNzJYsibsEAAAA9Q7BN1K9/LKbEnwDAAAEjuAbqfr0cdPFi+MtBwAAQD1E8I1U1dVuSs03AABA4Ai+kapBA6lJE5f1BAAAAIEi+EZNK1dK06fHXQoAAIB6h+AbNR1wgAvAAQAAECiCb9S04YbSZ59JS5fGXRIAAIB6heAbNW28sZsy0A4AAECgCL5R0y67uOny5fGWAwAAoJ4h+EZNzZq56bJl8ZYDAACgniH4Rk1VVW5K8A0AABAogm/U5NV80+wEAAAgUATfqMkLvhcsiLccAAAA9QzBN2pq2tRNjzsu3nIAAADUMwTfqMmYuEsAAABQLxF8R6Vly7hLkL+GDeMuAQAAQL3UKO4CVITPPpPat4+7FPnr2tVN+/SJtxwAAAD1DDXfUdhhh7oVfBsjbbut1KZN3CUBAKB8ff45yQlQMIJvZNaihbRkSdylAACgPFkr7bijdMABcZcEdQzNTpBZ8+bSF1/EXQoAAMrPhAnSd9+5+598Is2ZI22wAQkLkBdqvpHZokXSzz/HXQoAAMrPNttIv/1t8nHbttKZZ8ZXHtQpBN/I7De/cdM1a+ItBwAAdcHjj8ddAtQRBN/IbL313HTp0njLAQBAufjsM2nDDeMuBeo4gm9k1qKFmy5eHG85AAAoF599Js2aJe2xh3t8553Sv/+dfN7aeMqFOoXgG5k1b+6mZDwBAED673+le+919195xQXal1wiHX20dPvtbv7DD8dWPNQdBN/IjJpvAKjpu++km26ihrPSTJvmOlh+9pl77DXN9HijWJ9zjsv9XRdMnChVVUk//hh3SSoOwTcyq65201tuibccAFBOjj1WuuYaacqUuEuCKC1alLz/+9/XTCnoNUORXFAbpxdekPbeW3r11dzL7b67tGKFmyJSBN/IrEkTN3322XjLAQDl5H//c9MVK+ItB6L1j38k72fqcNm6dfK+Vzsel+efl955Rxo2LPdy8+a56fffh18mpCD4Rmb+C0ltFi6U/v53/oYFUP95GaC++ireciBas2a56WOPSeeeW/P5TTaRrrzS3W/YMLJi1bB2bbKso0fnt461ria/b19p5crwyoZ1CL6RWceO+S/bqpV08cXSo4+GVx7Ub19/LV1/vdStm/TUU3GXBsjOa9u7fHm85UC0XntN+tWvpNNPd9956YyRbr7ZVVwNGhRPEDt1qgv8R4xwjxcuzL18166pj196yW0DoSP4RmatW0udOkmnnpr/Ol9+Kf3wg7R6dWjFQj119dXSwIHSN99Ip5wizZ8fd4mAmm6+Odn298EH4y0LorVqldQgj5Bpww3dD7P33w+/TOm+/jr1sdesJJPnnnPX23QkWYgEwTeya9689lSDM2cm7y9aJG2+ueuMtGpVsiYAqM3s2amPa6uxAeIwaFDy/owZ8ZUD0frqK9fG3xv5OZfnnnPTMWPCLVMmY8emPl61KntFxvHHpz4+5hg39XcsRWgIvpFddXXtI1z6e/x7nVAGD06ud8MN4ZQN9ccPP0hvv506L1ONDBCXJ55wlQn+H4nTprkAywvM6PNSf912m5tut13ty26yiZv6f6hFxV/TfdhhbvrzzzWX8wfYXbq4c/eKK9xjar4jQfCN7PIJvpctS9730hLOmSOtv767T7tI1GbSJDf961+lI49096n5Rjk5/XQ37dbNTXv3dtOdd5a23dblSv797+MpG8I1bZr0yy/S1lu7Dom1advW5QNftSr0oqV45hkX8G+8sQumvU6hjz0m3Xiju3/DDdKBB0r9+yfX82q8vb4MV18dWZErGcE3squudjWS06dnX8YLzgcMkK66KjkyJpCPpUtd0C25L6y77nL3hw+Pq0RAdnfe6QKbjz9OzessJUc+RP3xww/SFltI//d/0gYb5L/eTju5f0kefzy8svlZK514oru/Zo2bdurkprfeKv35z66i7LrrpJEjpVGj3HNffuk6ukuu82XDhtR8R4TgG9lNm+amJ5+cfRkv+L74Yjfq2+LF7gPuadvWTb/80vX+/ukn/p4N07ffSnff7drqP/igdNBB7u/y3XaTevaUnnwy7hKm+sMfpNdfd/c7dZI228zd5wsA5WirrdzUmJrnqNfRfMkSOgzXF/6RHzt3zn89r2Y5qnbf/r5ZZ5/tpumZTLbfPnn/l1/c98G22ybH9GjQwI3O+e23fEdHgOAb2XmDSLz7rgusM/GCb29ETMn9FStJ22zjvqCmTnUf/F/9yrWHe/758Mpc6a64QrrkEpcW69xzkymnPvpIGjdOeuWVWItXw5AhbnrCCa6pUoMG0p57uhHinn8+WYsDxMVrR9u3r2sf6xk3LvPyv/lNslkK6jZ/84ytt85/vU6d3C2qjCfedV5K/kho1Ch1mfQRWb/4ouZ2vGwu330XXNmQEcE3slu7Nnn/mmsyB0Jem29/8H3IIa7W/IQTXJtvr/bA64n96afFl2nVqtR25n7WVnZP7XfeSf6wmTs39bkNNnB/hT77rPub0V+jE4dFi6QLLkj2CfDnld9yS9eJ7Zhj3I8GIE5ec4P27VPnZ6sdfPddhp6vL7zvk759pSOOKGzdZctyp/oL0i+/uOkf/iAdfXT25bbZxv1rk82++7rpggXBlQ0ZEXwjO3/wLbn8y+m8mu9mzVLnd+yY7MCx556pz+X68NfmqKNcrdLJJ0u33+46dxrjbiefLK23njRhglv23nul444rfl91yZQp0t57Jx97fz16zjsv+UUwcGD8zU/++9/U4ZovvTR5/5FHkm2+o/ryAmpz7bWpj3/3u5rLMNBY/bFsmatEOvlkdz3q3r2w9Y880v3rW1vSgiB47bZvvNF9B3o23jh1uTZtkm3DMzWjadHCTSu5EisiBN/IrmnT1MfPPFNzmWzBt+QC5QsuqDk/n4EKsnn5ZVcr+vTTronFu+8mnxs61E2//dZNL7pI+te/KmPQHy8F2umnp7btO+II6cMPXS93/4hrcWcT+fjj5P2XXpI22ij5uFGjZHvFQw91nYVQN/z4o/Tqq8nhreuq1aulXXdNre32Ush57r7bVVB8911yqHl/ikzazdZtXp+n9AA2X975cuGFgRQnp+XL3f6qqlLnv/6669Du9cMyxqXNfPxx972Qzqswi6rPzYIF8X8XxYTgG9ndcUfNeempA5cudR020tuXSVKHDsFmAMjU3OSf/6w579BDpdNOSz5+7LHgylCuXnjBTc86K7VG+cYbXRDRoIHrte+Ju2bDP4T8ppvWfN6rgZGiyxiA0u2/v2t25v8xFbZffnH/ynk/uoMwe7b7gegfRCzTP3bGuOvc1lu7zuW33pp87o03gisPomVtMgvTr39d3Da8f/PCbuLnDaRzxhk1n9tmG5eFzGtO0rixy2hy2mnJZAh+Xq35b38bfqXVAw+4fj6tWlVkPzCCb2R32GEutZZfeiC7bFlqe+9Mdt89tWY8W5vt2nz/fc15//538n6rVsn7/mYV6U0w6iPvS795c2mHHVy7vxNOSGZnkFxzjgcfdLV5cQffq1a5DCxffCHtuGPN5/3B9/Tpyc6/fm+9Jd13HzWM5cRr8hXlezJsmPsxd999wW1z5MjUx/m0923TJvVxFDWeCMcPPyQrdrzc7oVq0ULab7/wa5G9jr+Z/n327LGHdP757jsgF3/TmrCb/P3vf8nY4euvw91XGSL4Rm7pgWt6L+jBg2tPq/Xeey5Q8hTTBm7EiOQIXOkmTXKdQefPTw02/cIa8GDWLOk//3F/O6e3kY9LixbuR8nQoalNh7p2dfnYW7dO/mj54AN3wf3883DKcuCBrnbQG0jHs2yZy37To0fmGsWWLVODmUw1I/vt5wKciRODLXOcpk93X0r+JkJ1hf8zLrn39cUXw9/v//7npl9+Gdw2vVrPrbd2/+zttlvt63gDmUjSqafWrXSDK1a4wVWWLHHX0ltuif8Hepy8phDPPpv6j2GhWrYMP/j23qdf/Sr7Mo0auR+n/o7tmTRsmLyfqVlKqZ591qUzXLPGNRndaCP3HRVVSsYyQvCN3Jo3d7VY3l9a/i+URYvyC6SNSa2VfuCBwstxxx1uoANvOyeckHxuyy2T7cgPPDDz+mFdAP/4R+nww12+VG8I4qj5axkzNeFI5/+78bnnXPAa1l/kXg2iv423te68yTUgU6NGbljkOXPcY//f/+m8Zeq6CRPcl+N227m8+XWNfzyA1q3dtNAMEcXwKgQypU4r1rx50rHHJoeO/+Mfa1+nXTs33XJL10545kx37cvUUb3cPPSQ+8Fx++3uB9NVV+X3musr7/vCawNdrBYt3HmZ/sM0SF7wXWpZPd520itMgnDCCe5cmzHD/ahds8Z9virw30uCb+Tn0UfdL1T/h6SQoKdbN9eExRi3jUI/bJ995tqtzZ/vLoyPP+46ET7ySGoHzjvvTM1KcP/9bhpWLc5PPyXvv/qqm86c6Wr7P/3UXVgeecTl3Q7rAuN9UVx4YX5tbfv0cTVcw4YlvxR++CGcsnm8EdUkV+ttbe3NlRo1cm0CJTd0t79jjr+j7b/+lX85vBHd0psVlAP/v0oTJ7pMCeXyb0o+vM/C9de7890zebL7kTdsmGs6lp4GsxBr16YekxUrkn+Pr11be4o0a5OdkyXXidyY1PP/3XfdZzg9tWBt9tzTvc6XX3a1e97w3v7+DeXK+9dg1apkv54JEyq2M9y6a6q/+VsxvB+h++1X2nZyCeqHgserYLvsMunhh4PZ5o8/usozz1NPuQqZXXd1nxt/k9KVK933fX0PyK21FXPr1auXRQm23NLaDTe09r77rD37bGvvvtsLo/Pfxu23u+V/+snaV1+1duhQa+fOzb3O99+7dQ48ML99vPqqW3733a195hl3/5138i9jIZI/JdxtxAhrd9gh+bhz5+T9MWOC3/+MGe51StY+/HB+69xyi1u+qipZtiOOCL5sK1cmt3/QQcn5v/zi5v397/lt5/DD3fJffJGc9+CDyW3/7nf5l+mEE9w6Xbrkv04UVq+2tk0bV7bGjZOv7S9/ibtk+TvwQFfmCROs/fHHmp8N79aggbWTJxe3jx13TG7n3/+2dqutUrfdqJG106ZlXnfgwNR1rXXnpeSuGR7v3Prww+LK6Jk2Lbm/jz5y855/vvTtBs3/Xt10k7XPPZd6TCuRd658+mlp25k4Mfzj2L+/2/6PPwa3zc02C7bc++2X+Vpw2mnJ68Znn7llb7opmM9fmZA0xmaIR6n5Rv7WW881FbjgAveL2OvF7a/VrI1XM/u730kHH+xyjl50Ue51fv7ZTf1NTXL59a9d+YYMSfbeDqMGx2b4Zf7qq+4vtV12cY/9g23kajpRrJtvdu22JddcIR9ebc7y5e5fgu23z9yhsVT+Djv+1HPegBC5mp34eU0wXn7Z1Zob42oWJVcz/o9/5N+syKt1LbemKr/84sq0/vrJ9saSawNeV2y5pWt60b27a3aRrexr12buPJ2Pzz5L3v/zn91f44ce6mqrL7vMZWjItu2BA5P3p01zx9yrof/lF9en5Nxzkx0l8/08ZeNPTbjrrm569NHJzBPlwl/r36pVzcHUMl3n6juv9r/YzpYef7+V0aNL21Y23vtT6D81uWTqP2Ft8SMO//RT8l9MvyuuSF7Lp0xx1/d33nGPve/9eorgG/nbbbfUIOfWW10HjT598t/Gsce6qb+T3FNPuWYaS5ZkXsfbZ74dX5o0ccMCd+vm0oD5txGEyZNd++7zz6/53F13uQAv0xf3IYcE3/zF33s9n05hUmr7+3btpA03dH8J7rxz5o6pn3/uMt8UcgxvvTW1Ccynn7rg7IorXDpEqWZ2iGy23dZNP/oo2bnOM3++C7h+//vat/PQQ8lhmOfNc52FFy1ymXGiGAjDb9Uqd654HSu9VJH/+EeyuYIUXb7dIDzwQGrGhc03rzlWgKfQ1/XWWzU75nq5tQ880GVz8Eb2GzZMGjTIBRDeNSU9gLzmmtTz84EHXHvnBx9MfgZqaxZVm8aNU9N+eoFLsdmeCvXvf7uKiGydd1etcsfUqyiQ3PFKz3AVZArHuuD7791ns23b0s8B/zXuv/910/R0vaVYs8Z9B2y9dWmD16Xzf0f4x81o377w5AVvveWaMG25Zc3n2rd3g+ZJbgTstm2Tx6mUkbDrAIJv5K9Fi5pfmoW2M6uqcgG7lx7Js+eebvv+jnkeL2Atpv2dt06QQcxFF7nOSP6Oo/4LX+fO0vHHp67jZWFJzxZTivnzk1/kBx2U/3qHH568f8YZyZq4MWNcrX26P//Z1Tpnem8yWbFCuvLK5GMvBdzUqS7AWbrUDdl98MH5bW+jjdzonS+9lJrq0p/9xqtNz+WTT9zUy/QyfrwLtk47LXO++DA9+KDLA/z3v7vH3nSbbdw/Al4H57qSccJrh53e4fe551Ife7Vz2X5oZzNkSPbn9tnHTb3P+p13ulrw7bdPHsf0f53SAyDv3PD06xdMMOPv3PzKK6VvrxDnnCO9+Wbmz7SU+u+P9+/jkiXJc+6YY9y0ngdBNdx5pzs/dtghmO116eKmt9zifhA2a1Zc0oFMvMoIf7AclKuvdlPvH+777nP9JQr91/C999z0jjvcv0o77SRdcom7Drdpk/yHaNas6CtBYkTwjfy1bFmz1qaYHNr+YdDT+VPKffWVq6HygsVyCb7T/5IbNCh1+3/5i3TAAcnHxx6bDEAnTHCj4N18s7tfSi2IvylHepCTi3/44ZYt3YAKnky98r0vX3/n0mxWrqz5A+PQQ1Mfjx/vAu/GjfMrr5R8H/25nPfd19WuSrWntHvtNVfT3b27qyHddFP3nnlBWa6/OOfOdbXuQfb+9/Jhe81gli1z6el69nRB35Ah0l57uc9Av35utNhrr01+kUXl+efdZ/y662p+6S5d6v5xGDHCfU6lZMDmOfzw1FaeH33k5i9e7Jp+pA+gNGWK+6fl3ntd4Ojfl+eSS6TzznP3DzssOaJrpuuD90PL60Q2dGjm2sz0v9ODaqZ2xBHJHPyTJwezzcWL3fsybJj7ETd2bOblvHPryCNd0NOhg/uBvWKFa8rlf69uvtkFhR9/7D4L/fq52l/JHbO77gq2xraczZrlfiT6Ow2Xwp9Jyut8m+vHZD7mzHHvu3f98qe5DErfvm66116pP0S9kaQPO8z927rzzsmmj5ksXuz+jd5nH+mee1y577wz+U+V17F+8ODU9W64obTO2eUuU0Pw+nqjw2WJbr21ZoeJu+4qfDuPP+7WPf10a5s3T93e9tsnl/M6kni3efMK39fy5cn158wpfP1MmjVLLZfXect7/MEH7vEJJ1jbsKE7bv/7X/YOaPPn177PX36x9rbbrF271j2eP7+0TlG33uo6tljrOrzm2lbLlm7+uefm3uby5da2bl3z9c2ZU3PeOecUVt4PP0yu+/nn1i5c6OZ/803tx2D27OQynTq5ed26WXv88db+4Q/J55Yvz7z+Hnskl1m5srByZ9O+fXKbXkfLCy5IXea882oetz59gtl/vjbfPPu54XVs9m5Nm1r7+uu5t+edC3ffbe2++7r733+ffD799Waa/+9/W/u3v7n7/fsnl0k/j73bRhu5zr2StS++WPP5Dz5I3t9uOze98srSj51n4UK3Tf97/uOP1p54ojv/CjVoUGr5N9ww83Lp1ynJ2t69rX3hhZrz1661duutk49vusnaFStSO9698kpJh6FOGDzYvdZddw12uxMmpB7v5s2L39batclrRs+eqd85QVq1KrXM1dVuuvXW7vlsn9V0F1xg7QYb5N5X+nXGuz3/fHCvJyaiwyVK5v0SltzfZm+8kbndc2322Ufq1cvVXqXXAPiHtJ09O7V2NFOHjdo0bZrM/R1UzaW/7fnIkclcxlOnSk8/nexcNXSoez1XXJE6cli6fIYfPvtstx2v6Ye/U9nddxdUfEluW3/6k7vfunXq9u6/39XErlzpahy9Gq/a/j2YOze1k+WoUa4Wyd+xslEjNy30X4yePZP3e/RINnfytyNs1MiluUtvk+hPLXfmmcn9//JL6nPZRnTz1zbPmxdMDaDXFvTss5M1yunH5L77Uptn9O3rli2201O+hg93Nas//JA5BeXSpa7W2n8+bLGFOy61DcXtnQvvv5/8l+Wnn9xxzWc469mzXY3Z+ee7tqH+4dxbt3afx/QmRDNnJmuHO3eu+Y+Lv6/Eqae66YYb1l6WfLVo4Wrre/VKzrvsMnd9+NvfCt9eehOrbE2uMuX8/+67muf5q6+6ms0xY9zz06e7PN9NmrjH3r80+fzzVdd5zY/uvTfY7aZ3hiylM+/SpclrhjcyZKkpETNp1MiNTu15/XX3j8iECdmbMqVbvdo1Y6qtc31681Xvn7S61OG8UJki8vp6o+a7RDNnul+j3bqFs/0TT0ymgEtPVda3b/Hbff99t40777T2pZesfeyxZA1yPkaOdNuwNjV1lFdjlC+vBjn9dt99ta+7zz52XY2htalp04JIMbV0aWqZnnyyZhq3XLUb559v7d57Z1527dqa27n88sLLmK2WL33bQ4YknzvtNGs33rjmsfbSM/pv221n7aWXWrt4cXL96dMzv2deDWwxrrjCrX/qqcm0i5K1N96Y/XVL1vbr56Zt2iRr/oPm/4dm000zv/+ZjkW+NYVr17r3MNM2fv3rmvMmTkw9f/KV7T2bOTM1lePBB7vlH3nEpeBcvtzam292tb5hyHTevf12fuuuWJE8lzfYIHUb771Xc3n/ee+/3Xtv6mPv2pbNggVuuUJSetZVW2wR3vfbkUcmj/nuuxe27pIl1l53nbUXXphMl+q/TZ0aRoldKmBvH+PGWXvNNdk/W1tt5b4rPePGJZ/r2jX3frx/wTp2dNP0f9XCShUcAWWp+a4xoz7fCL5LtHat+2IaOzac7ffr577wrU3+rSy5JgelNBmZNKnmheJ//8t/ff8Xv5c3fL/93N+/hXj8cfe37y67pOYrPvnk2tc97jibEnzuuad7fNBB1q5ZU1g5svECfKnmX9u/+pW1xmT+sfHtt6nLduiQbNLiuekmF6Bvs41bxn+RLtWQIan7/+tfk8+1auX+Jj3gAGtvuCGZU/6MM7J/ifjL5r3fmW4dOxZXXm/98893+b232879pfvyy5mX//pra7/6yjWx8b54v/mmuH3X5pNPar7Ogw920z33zPxD6pZbXBnzNW+e+zHt30bTptZusom166+fOv/FF92PIcl9QefrqKPcOqecktzWhhu64+0F35MmBffZyddLL2U+l/IxY0ZyeX8TEclVKKRr0SL5/BNPJO/feKObDh5s7T//6Y5JbSRrf//7wl5rXbT55i6IDIO/iZu/eWU+RozIfh2SXBORsHj7+PZb96Pf/wPy6aezn8v+fPG1VZ79+KO1r71m7aJFyeZ/6deYOorg2xJ8l70LL3Rthq11tQ/el3Kp/O2+vdvHH7vnli1zX0KPP566jleTMHZs6kXlwgvd/RkzSi+XtW7gIsl9CT71VGqtq99ll7nlqqvd4222Ke3fgEzuuCP5Wo85Jnl/m22S7f27dHGB3+zZ7pitXOm++P3H9rnnsu/j0EPdMkG2URw/vub7e999rp1vw4bWXn11zXUeeCC57MYbp7Y5vO8+a//zH/cavTag/rblhQZNnjffTG27fsUVhb9Wr73upZcW9q9Lbf7+d2u7d8/8GgcPdv9qdO+eefCcYvpipLdn79Ileb9fPzcIj2TtVVcla8Hy+YcoE2+7I0a4x17wvWhRcdsr1dFH1zyGY8daO2pU9nWeeMLaiy9OLu8F0N5to43ccmvWuBrqjz9OfX75cmvvucfdX289Ny3kh0f79m5gNb/Fi6096yz3o7aumzzZXYOl8Gr4n3wy+X40aVLY5/ff/655znjvp9cGOyzed/Evv7jHd92VLMOoUZmviatWWduuXXLe/fcXvl//P2/durnz9eefg73uRYDg2xJ8l70rrnDBtlfTJVn7xhvBbDv9AuF90T36aHLekiVu3htv2HW/6r2OLd5FZeedk19mYZTr5pszL+evKR892k1POCGYMni+/trVyEjW9uqV3N/gwamd0lq2TP4IyVST9+ab2ffx888ucAj6Arp8uasd/tOf8jumY8cm/5b/179qNpmRXGe+O+909zN1GvUH36tXu5FNs3VAfuWVmus+8EDhr3PkyOT6r71W+PrZZHpt3m3iRPdvTZs2mTssFuPrr1MDcP/5/cEH7rMouR+BXsdrb5TIYl+b92/Xyy+7f67i+hL3auUz3TJ16F271gVr/uWee879SPF+vEuu6Zg3GrB33dprL1f7v3ZtakVCs2aFlblLF9cs0O/tt0s7B8qJ/9j6/zkLUnoAne+IxNbWrOCQrH33XTcNsnNwJp9+6ipfvM/Ll18my7BqVWow7p0L3o9nyTXlKWak0K+/Tt3uX/7ipk88Edxri0C24JsOlygf1dUuDZaXFmzQIGm//YLdh9fx0ess5k8x9+GHrjPb+PHu8VtvJTu2eCntlixxKZayDR5SqmwDWjRsmLzvdQDs1CnYfXfr5vKvH3NM8hjceacbFGe33ZIjTS5alMy37S/v6NEuPZeXdzmTjTZynWmCHBBCcu/H1ltLN90k/eY3qc9l6oy0006uo6u17vW+/LLLmesfJOaXX5I5j/0dgl56yR2TjTdOzps3z6XIu+SSzOXzj3QquRRx3shuhfC/lnxym/utWuVSgr37rhtU6f/+z3VazLWdpk2lrl2lHXd0nxmvg9fpp7vpZpsVVgZPt25uABrv8+gN1LXddu5cq652g7+89577XHbrljoYTCF+/NFtZ5tt3ONDDnGdxYM+B/PlpfY86CDXSXTnnZPPZUpx6HV+vuUWdz7Om+fSlz7yiDuv7r/fLXfeecnRhr3r1mWXuUGkjHHnfP/+bn6hr715c5dS8qWXkvOmTk3e//579xnKpzPwqlWuY+yUKS79orWFlSVs3nUuaIcd5kZ29gaHmjYt/3W976tx41zK0TfecANLTZ7sUvKFaccdXQd975zxBj2TXKfMiy+uOUq1l2522DD3HbHjjoXvt1s3d26MHOkef/ihm9aTvPME3ygfXuDjZaDwD1BRKi94vuoqN73wQncx8TJ+SC5Tw9ZbJwOohx9OPvf9924krq++Sh0Zr1SdO6c+ztYrfMkSaf/93X0vp7eX3zhojRsnR8Vr0iQ5v2PH5H3vy8AfbPbu7TLLxBXUePzDekv5jaTZsqULzvw/aF56yQUukjsmXuC70UZukKCffkoORV7bYDjeIDqeYt87/2AaXr7dfD3wgPSrX7m8vQcf7G6//W3NTDxVVcnPoncetGvnfhj/8Y9Sgwbui18qfQh2L9OI9+W8117J5zbd1B3jL78s7TO38cbudZcLL8/4rru6c8o/wM+RR9Zc3suI0qaNWz4965N3Lj3+eDJbi+Q+h+nXl3bt3LTQwUyWLHEBfd++ybEe/EHQMce44PL992vf1vDh7jrRpYsbfKxBg2R2izik//gsdVTLbJo0cZmptt7avYf+HzK5LFqUvM506eKCba9SKlP2nijsvXfqQGr+gLxx42TWoyCysHjjUrz8spsOH176NssAwTfKh3fR8y6GQaZPuvFG90vaG+AnvdbBCyzT0xHuv39y4B8vXdkf/hBcuUaPlp54Ipku0D8Ajt+SJS6I2HlnV+PUvr1LuRcG/wBB/rR6F1wgPfNM8rF/cJ6jjw7vS6tQd9zhauDvuEP6178yBzTZXHZZ8n7Tpu6LxPvh9sknrvZv552TKTa9c6K2NIxeikWPN6Ryobp3T9YypW8zk9WrkwG0P1XcOee4mqQttkimnpszx9U0LVvmalal5KBGhxySXPe991yKxEWLkl+IxbrjDvf5OvVUl9bwnnuSz3k1upL06KOl7aecHH+8e21XX50cKMurkfZq9/y8f5fSBzDyeD+E/H74wVVipP84uvzy4srs38fEie5c8QetXgrUfAZFyZSmrtRBZ0rhpVKUXKrYKFRX5/f5lZLl69XL/TAuB6NHu4GZPAMGJP8N9qcMDeI7PH0b3mi6dV2mtij19Uab7zL3yCOuTdc//uGmXgepoDVqlNqWbNNNU9vU3XhjMuuC1zbce+6gg8Ipk7XJQTE23NB1LvO//jZtXBvZKLz1VvL13n57zee9gUhmzUouF3XWiLCsXFmzbWW/fpmXPfZY16lo4EB3XnjL//e/yWXGj3dtyv3bu+OO0su5zz7JzsnZrFiRTEn32WfWXnRRsgxeykr/wC/+NtDXXuvmvfSSezxvXjzte+tLm+JsvLbuH3+cbNP61VfuuZtuSr7+Hj1yb2eTTWq2u81kxQr3fKGDvPgzdeS63X23axueqxNupsF/JNf/ZuBA1/H8u+/c94CXiWXiRNfh00ux+eGHroOk35Il1j77rLv/+99b++CD+b223XZLfkai4mXh8bIv+U2Z4joaP/OMS0fqZTjKlE6ynFx9dc33dNy40rc7bVrqNtu2LX2bERIdLgm+y56XU/TKK900jFG7rE2O1OW/+bM4XHhhMlCZNs2tE2TglE2mLyVPVVVxo+EVwz8ipH/0Qc+YMe7LYe1a92VZ3z5Xv/51smOt5HKFZ3LLLdmDEE+m54I4r/v0yT6yoeeHH5L7fOaZZHrFY45Jpir08uqmZ86ZP99lKPBSmK1d60YE/eMfSy97IR58MNzPXNw++cT9kFq2LJnJxvux5z9nrr0293bGj3cBej4/Vm66yS1fiO++c7nc08/ls89OfeyNcJueatTPS4F45pnZPz9eKkovK9Xpp9uUH4PemAn+H4wDBiQ/X/n+aPvuu+SyCxYUdkxK4Y3TMGxYzeeOOCL1WDRs6LLZzJwZXfmK4U8tKFl7yCHBpEBcscKNTCy5Y1NVVfo2I0TwbQm+y97w4e6U9GpWgxg8JhMvpZo/17O1ycEnzjvPXdT9F+NMtYNBy5QFwVpX+yO5QRYQHa/2N1fgs3atuy1f7tKu+c8Tb1hx7zZ6dHDnjzdQT67t+VObdejgBrpIT0vmDWzjr61HfLbYwv2jkj6YVz65uL1/rLbdNpyyvfZaapl23bVmwOXdNtnE/Rv28MMu84XHu8b6r2X+wWfSb96Q9l5Wl7POsvaLL5LPf/ihy/KyZk3yun7VVcnnvQxW2fhTf0Zp6lS3z+uvd0H19OnJ1JfZjkFd4JV51qxwtn/DDW77YeY1D1i24Js23ygfXue+L7900zCGzJWSnZC8jnhep0Gvk8hxx7nOSv7210cc4TrlhdmZMFP2EmOS7W9rG6IXwfI6luXKBmKMuzVt6jqPeRo0qNl+v0WL4M4fL/tNpuHfPePGJe9Pn+7aDu+5Z+oyd9zh2vOmz0c82rRx/RS6dUud7892lM0OO0iHHirdemsoRVPPnqkdYqVkBpl0P/4o/ec/rm/AmWe6eQsXJjsO9uqVXHbTTbPv85JLXL8Tr937I4+4cni8zDi//nUyE4+/LXJtbbi9Dr9Ra9PGvafXXec6E3fo4PoXWVtz2XJp552Pk09202x9l0rlxQReXxTJdcBs1y7Z2fitt9x3pTGuU/Ommyb77ZQRgm+Uj1WrUh+HFWyeeabL9HD00a4z2qBBbn737u7ilylV3tChNdPFBe3WW13A9MQTyQ59kvTqq5m/+BAuL2tKvmklM52vrVsn7wf5Y3Knndw0Vwe3L790HXO9TsbnnCM99FDqMj17utSD/LArD15KywYNpGefdekrJ07Mb93113cBr79zbJA22kh6+21p8GD32BiX5cKr79xyy9TlvZR6X3zhOrh7r+ORR1xmFM/f/pb9B8Pq1ckOwelZjPy8FIvp6Si9lIuZ+L9v/BlnotCyZc3MMB99lNrB3VOXOhg+8IBL2erPkhUkr/O8v2PyP/7hOhf/85/u8XvvJbP5fPyx+yFYhgi+UT68rAyeBiGdnkcfLb3yirT77i6bRT61kVVVyTRdYWnUyKUNO/XU1Lypm2zivoTjqqWpVF4e7y5d8lveC66PPz4577rrkveDDHC9dHNvv519mQUL3JfQhhu6x2Gfvyidl7LSWvcPXM+eLs96OenQwU132CF1fnq6TS+N68KFLquOFxinp5Bt2tSlHvTbfXeXXWnq1GSK1b59ay+b//MmuXR43o+FdF72n7//vfjsQ6XIlLfe+5fAL9+sKOWgRYvs/4YEoVMn91599pl7PG2a9M47yeeHDMmc7vLSS8MrU5Hq0LuKeu/gg93fcXPmSLffHndp4uVvguIfzALROf5492Vy8MH5Lb/PPtJRR7mBNFq0cH959unjAuR27VIH5SmVNzDLihXZl1m50n3B3367a47gr21EeRozxk0zNT8oFwcc4JoxpddEt2vnUieed5770bBggcsLny490JZS06bedJNLq/iPf7jHX33lAu/LL3c/JufMcT+Iu3Vz+2re3OXdnzvXBe1vvulqy48+2q0/cmQylaOfl/IwzhSp55zjBtzyPPusm+6yi3stP/9cXjnqy8GSJcnr3m9+k/pvQb9+NZdv3Lgs/zU2tpw/5AHr3bu3HeNd3IC6wKuVr6DPKfJgbfKfoWXL3D8zb7/tAuw1a1zTpWuucc0C/vWveMuK/Pn/hatrn/njjnPn2tNPSyee6OY9+6x0wgnJZZo3z54TP/1a9+CD0rnnutFks9Ve5+I/lqNHJ5tfSa4p3znnuD4T77/vAt043XOP+9HumTKlZjMeOOee60bO/PHH5ABDv/2tdNttyWZbbdq462CTJqmjFsfAGDPWWlvjrxWanQBAXeMPLF580U3HjXN/8S9Z4v6WXbw4vE7LCIf3j5/X+bsuufpq15ndX7M9f37y/sEHp9bypuvb19Vke7za8GKPhb/te58+qc+dckqys7J/5N64HH546mOvqRhq2mAD9w+Lf2TPDh3cgFIdO7pbixZuNOCYA+9caHYClLOhQ2tmPgD8JkxwHdO8GsUmTdz977+nI2Vd84c/uKZL7dvHXZLC9ewpTZ6cOs9rEvC73yVH8c0mfdjw3Xd3tZvFHouXX3aB+5tvusdffOHaz3/6abITp5S7I2dUOnZ05T30UPeYH83Z9e/vMsX85S/u8cEHZ25WVOYIvoFy5v/LFvDbYQfXEff6612wXVXlmqK0aSPNnOmW8Xr9o+6oT80NvOHti01lWWo/CS+drOQ+L95nxpPe+TNOdIjOz5ZbSjfemAy+X3kl3vIUieAbAOqiDz9M5gC+/35pwABXY9aihUtbJmXOqABE5de/dhkp4mra8fe/u/R01roc1P7A+5prXDv1crH99nGXoG654YZwx90IGR0uAaCuSv/yadvWZcrxrnO33eayRACVbPXq1DbCkvuM+Af7KQd0sK93snW4pOYbAOqq665zzU48m2wi3XWXq3FcsSI5SidQyRo1cqNrfvmlG6BqvfWSA1WVk/79a3a+RL1E8A0AddXAga6j1tix7vH++7vh4q+6yj0XZw5joJzcc0/cJajdww/HXQJEhOAbAOoyf97knj3d9LLL3IiDF1wQT5kAAFmR5xsA6rILL3TTU06RTjvN3W/RQvrb38o6zy0AVCqCbwCoyzp1ctP6lKIOAOoxmp0AQF3229+64bePPTbukgAA8kDwDQB1WcOG0llnxV0KAECeaHYCAAAARITgGwAAAIgIwTcAAAAQEYJvAAAAICIE3wAAAEBECL4BAACAiBB8AwAAABEh+AYAAAAiQvANAAAARITgGwAAAIgIwTcAAAAQEYJvAAAAICIE3wAAAEBECL4BAACAiBB8AwAAABEh+AYAAAAiQvANAAAARITgGwAAAIgIwTcAAAAQEYJvAAAAICIE3wAAAEBE6lTwbYxpYIy51BjztTFmuTHme2PMHcaY5nGXDQAAAKhNnQq+Jd0paZCkryRdJOlfkn4n6T/GmLr2WgAAAFBhGsVdgHwZY7aVC7iHWWuP9s2fKunvkk6Q9HRMxQMAAABqVZdqi0+UZCTdlTb/YUlLJZ0SdYEAAACAQtSZmm9JO0taK+lj/0xr7XJjzOeJ58vO/FWrdM7XX+u5OXMkSU0krUxbpnWDBqpu0EA/rl6tRpKqJC0qYB/NJFlJyxPrLi+92AAAAHWekXRv5846f/PN4y7KOnWp5nsTSbOttSsyPDdDUltjTJP0J4wxA4wxY4wxY2bNmhV6IdO9NGfOusBbqhl4S9K8tWs1Y/VqWUmrVFjgLUnLlAy4CbwBAAAcK+mSKVPiLkaKuhR8V0vKFHhLyZizOv0Ja+1D1tre1tre7dq1C61w2Rzepo2Oa9Nm3eMavw7kar43bdRIRlJjSS0L3EczuRpv+aYAAACVzki6q3PnuIuRoi41O1kqacMsz1X5likr6zdurGe3317Pxl0QAAAAxK4u1Xz/KNe0pGmG5zaVa5KSqVUHAAAAUBbqUvD9iVx5d/HPNMZUSdpB0pgYygQAAADkrS4F388q0W4+bf7Zcm29n4q6QAAAAEAh6kybb2vteGPMfZIuNMYMk/SqpK3lRrgcLQbYAQAAQJmrM8F3wiWSpkkaIOkQSbMl3SPpz9batfEVCwAAAKhdnQq+rbVrJN2RuAEAAAB1Sl1q8w0AAADUaQTfAAAAQEQIvgEAAICIEHwDAAAAETHW2rjLEBljzCxJ38Ww67ZymVmQH45X4ThmheF4FYbjVRiOV2E4XoXheBUmzuPV0VrbLn1mRQXfcTHGjLHW9o67HHUFx6twHLPCcLwKw/EqDMerMByvwnC8ClOOx4tmJwAAAEBECL4BAACAiBB8R+OhuAtQx3C8CscxKwzHqzAcr8JwvArD8SoMx6swZXe8aPMNAAAARISabwAAACAiBN8AAABARAi+AQAAgIgQfIfEGNPAGHOpMeZrY8xyY8z3xpg7jDHN4y5b0IwxVxlj/mWM+dYYY40x02pZfldjzOvGmEXGmIXGmP8aY3bIsuwmxpgnjDGzjDHLjDFjjDHHZlm2qTHmBmPMVGPMCmPMFGPMNcaYxqW/ymAYY7omyvhh4jUtMsZ8boy5OtO5YYzpZowZboyZZ4xZYox5xxizX5ZttzLG3GOMmZE45/5njDnPGGMyLFsnzs/E63/KGDPBGLPAGLM0UeZBxpiNsyxfsccrE2NMte+zeW+G5yv+mCWOTabb4gzLVvzxkiRjzAbGmL8ZYyYnyjvLGPOWMWavtOUq9novScaYgTnOL2uMWZW2fMWfX8aYFsaYPxljxifOm9nGmPeNMWekv7Y6e35Za7mFcJN0tyQraZiksyUNkrRK0puSGsRdvoBfq5U0R9JISXMlTcux7G6SlkuaIunSxG2KpEWStk9bdgNJ30paLOkGSQMkjUrs78wM2x6eeO4RSf0TUyvpsbiPka+MtyRe61OSLpJ0rqRnE+X8QlIz37KdE8d1pqSrJJ0v6bPEebR/2nabSPo48dygxDk3LLHdgXX1/JT060SZ/pp4/QMk3ZM4J36UtCHHq9Zj+LfEOWcl3Zv2HMfMrruGvS3plLTb8RyvjMero6SpkmbJXdP6yV3Lh0g6wbdcRV/vE+XskeG8OkXSbd77zfmVUs4Gkt6RtEbSo4nz4BJJHyXKf2t9OL9iP9D18SZpW0lrJT2fNv+ixJt3UtxlDPj1bum7/6VyB98fS1ooaVPfvE0T80akLetdnA7zzWuY2MYcSS188w9OLHtH2jbuSMz/VdzHKVGe3pJaZZj/l0Q5L/TNey5xAdrBN6+FpO8kTVQiW1Fi/vmJ9S9K2+7zklbKDXFbb85PSccmynoFxyvncdpJ0mpJv1fm4JtjZtcF34/lsRzHy5XrHUnfS9q4luUq+npfy7F5MFHWQzi/Usq0e6JMd6bNbyIXPM+vD+dX7Ae6Pt6UDKT2SptfJWmJpFfjLmOIrz1r8C2pS+K4PJLhuUcSF4f2vnk/SJqcYdlTE9s5zjfvn4l5m6ctu3li/j/iPja1HLftE+V8IPG4udwv+jcyLHttYtldfPPeTZxbVWnL7qWaQWqdPz8l7ZJ4DTdzvLIeo4aSxkp6WVInpQXfHLOUcllJj8l9wbfIsgzHy5Vpb/mCPkmNJVVnWI7rffZj2FzSArkfMA05v1LKdFCirJdneO5jSTPqw/lFm+9w7Cz3xn/sn2mtXS7p88Tzlch73R9keO5DSUZSL0kyrj3vpon5mZb1b8+7P8Na+71/wcTjH1X+x3yzxHRmYtpDUlNlP1ZS4jUZYxrI1XB+ljjH/D6Wu1ikH6s6dX4aY6qMMW2NMZsZYw6UqzWSpFcTU45XTZdK6i7pwizPc8xSHSNpqaRFxphfEm1pW/me53g5Byem040x/5G0TNISY8w3xphTfMtxvc/uWEnryf3bsiYxj/PL+VjSfElXGGOONcZ0MMZ0N8bcLHe+DEwsV6fPL4LvcGwiaba1dkWG52ZIamuMaRJxmcrBJonpjAzPefM2LWJZb/lMy3rLb5rludgZYxrK1WyslvR0YnYhr7+1pGaZlk2cg7NV81jVtfOzv1z70u8lvSZpfUmnWGvfSTzP8fIxxmwh6XpJN1hrp2VZjGOW9LHcl/oxkk6Xawd7oaR3jDEtEstwvJxuienDcu1oT5dr871S0pPGmDMTz3O9z+4suQD5Ud88zi9J1tp5kg6X6z/2nFyTmwmSLpB0tLX24cSidfr8alTqBpBRtaRMJ7nk/lbyllkZTXHKRnVimunYLE9bppBlvfu5jnl1lufKwV1y7dz+ZK2dmJgX1LHyli/kWHnLlNP5OVzS13LtH3eUuzi39T3P8Ur1gFz7yEE5luGYJVhrd02b9YQxZpykmyRdnJhyvJyWiekiSftaa1dKkjFmuNw591djzOPiep+RMaabpD3lmpdM9T3F+ZW0WK4J60uS3pf7kXeBpKeNMX2ttSNVx88var7DsVTu76NMqnzLVBrvNWc6NunHpZBlvfu5jnlZHm9jzI1yNWwPWWtv9j0V1LHyli/kWPm3WRastT9Ya1+31g631l4nV9t2mzHmqsQiHK+ExF//B0g6z1q7KseiHLPcbpcLRg5JPOZ4OcsS06Fe4C2tq7F8SVJ7udpxrveZnZWYDk6bz/klyRizvVzAPdJae7m19gVr7SNyP1h+lvRw4t/iOn1+EXyH40e5v3EyvXmbyv39Uy6/MKP0Y2Ka6S8bb96MIpb1ls/2V9Cmyv4XUmyMMQMlXSOXnuvctKcLef3z5L4QayybOAfbquaxqtPnp7V2nFwKrvMTszheWlf+QXJt4X82xnQxxnSRSw0nSa0S89YXxyynxA+XH5X8h4Xj5fyQmP6c4bmfEtPW4npfgzGmkaTT5LJrvJD2NOeXc6lcgPsv/0xr7VJJr8hdyzqpjp9fBN/h+ETu2O7in2mMqZK0g6QxMZSpHHySmO6e4bnd5NrAjZUka+1Pcif4blmWlVKP4yeSNjXGbO5fMPF4E5XZMU8E3tdJelxSf5voSu0zXu5vr2zHSkq8JmvtWkmfStoxw8V1F7mOJ+nHqj6cn83k/o6UOF6eZpLaydXWTvLdRiWePyXxuL84ZjklyrqZkp2gOV6O12lvswzPefN+Edf7TA6TtJGkf2Zof8355XhBb8MMzzXyTev2+VVquhRuGVPlbK/cOTVPibuMIb722vJ8fyKXg3MT37xNEvNeT1v2dmXPyzlPUkvf/EOUOy/nnnEfG1+Z/pwo0xPKMbiB3C//NZJ6+uZ5OV+/UWrO1wuUPefrKkmd6uL5KV+qqLT5+yaOzRu+eRwvl/btmAy38xJl/b/E464cs3VlapNlvnf98ads43i5Wu2FcjXg/tzIG8u11Z3om1fx1/u08r2cKN/2WZ7n/JLuTP/cJeavL1cjPVfJ9Ix19vyK/UDX15vcKHxWbjSp/ok3bZVcDVTZjCYV0Gs9Va75xDVytUTzfI9PTVv2V3K/7qfIjVp1SeL+Yv8FJ7FsG0nT5Dr2XC83ItVbieN6VoZy/Cfx3GC5dnWDE4+fjPsY+croXTC/k/v7MX3UswN8y3ZJXGhmSrpSydHOVks6KG27TeR+ja9KnGv9lRzt7Ma6en7K/TX7odwIl+fIdX57Qq4t7nylDkZR8ccrx3HspMyD7FT8MZP7sv8gcY6dK+kPctlObOLc8486W/HHK1HWAYmyfik3gNOVcte0lZIO9C1X0df7tPJukjhPPsqxTMWfX3LNSubI/Vh4MvGZ/JPciKpW0vn14fyK/UDX15vcL6rL5EalWiH3l8cgZRnAoS7flByiNdNtVIbld5f0RuIDskguddxOWba9aeIDOFuul/GnShvy2bdsldxgAtMSx/xbuRR+jeM+Rr4yPpbjWNU4XpK2lvSiXKC5VG5ghf2zbHt9SffK1Q6skPSVXGdOk2HZOnF+SjpOrrbo+8T7v0wu68k9kjpkWL6ij1eO49hJGYJvjpmVpL6Ja9CMxDm2RC738Z+UNoAJxyulvEfJ/ThZIncdHyFpjwzLVez1Pq28f0p8Bs+uZbmKP78kdZZrkvmD3A+EhZLelnRUfTm/TGInAAAAAEJGh0sAAAAgIgTfAAAAQEQIvgEAAICIEHwDAAAAESH4BgAAACJC8A0AAABEhOAbAAAAiAjBNwDUE8aYgcYYa4zpFNP+HzPGMHgEAORA8A0AETLG7JMIkLPdVsddxnJnjNkxcax+lXi8S+Jxj7jLBgC1aRR3AQCgQg2V9GqG+WtL2OZfJN0iNxxyfbaH3PDQYxKP95IbgvrL2EoEAHki+AaAeHxqrf1nkBu01q6WVAk157+SNMZauzLxeE9JH1lrS/nhAgCRoNkJAJQpY0ynRHOKgcaYE40x44wxy40x0xPzGqUtX6PNtzFmA2PMncaYKYl15xhjxhpjLk9bt5Ex5o/GmK98y71gjNk+Q7mqjDG3G2N+NMYsM8Z8bIw5MMfr2MoY86Qx5idjzEpjzLTE+s3zPA4NjDFtvZtc8P2F7/Eeksb5lmmcz3YBIA7UfANAPKoTgWO6ldbahWnzDpe0paT7JP2ceHydpI6SzqxlP/+StLekBySNk9RM0taS9pF0u2+5pyQdJ2mkpPsltZd0gaQPjDF7WWs/8y07VNIRkv4j6TVJnSUNkzQ1fefGmF6S3pQ0X9KDkmZI6inpd5L2MMb0sdauquU1dMiw7QsSN89liZsk7StpVC3bBIBYGGvpmA4AUTHG7CPprRyLvGKtPTSxbCe5oHOtpJ2ttZ8m5hu5YPcISbtbaz9MzB8oF5RvYa2dZoxpJRf03m+tPT9HmQ6QNELSc5JOsIkvBmNMT0ljJX1grd0rMe9AuYD7cWvtGb5tHCHpBUmy1hrf/C8kNU2Uf5Fv/pGJ13CmtfaxHMdDxpgquaYlknSYpPPlfoCsknS0pDMSj70vtLHW2nm5tgkAcaHmGwDi8ZBcrXS6WRnmjfQCb0my1lpjzG1ywfeRkj7Mso9lcp0vdzXGdLLWTsuy3JGJ6U3WVyNjrf3CGPMfSUcYY9pZa2cl9iml1prLWjvcGDNRUjdvXqLJSg+5HwRNjTFNfau8K2mJpAMlPZalXN62l0t6PbHNs+Tad/9f4vH5kt6x1o7MtQ0AKBcE3wAQj0nW2tfzXHZChnlfJaZbZlvJWrvSGHOJpLslTTXGfCXXBGS4tfYN36JbyNWuZ9rP/+QC7i3kfhhsmVj2myzl7OZ7vHVien3ilslG2covufbekjbwHko6QNKjiSY7Rq6JyT2+Jjxz6XgJoJwRfANAPWatfcAY86KkQyT1kXSMpAuNMc9aa08Iefde85M7JP03yzK1NQ/J1N778sTNc23iJrkfCdPyLyIARIvgGwDK39YZ5m2TmH5b28rW2p8kDZY02BjTUNKTkk40xtxhrf0ksY0Gif2My7IfLwD2lu0qVyueq5yTEtM1BdTyp/tZrrZbcj8cTlGy6cupcm29j01bHgDKFqkGAaD8HWCM2cl7kOhweUXi4fBsKxljqo0x1f551to1SgbYXnMObxtXJbbtrb+dXHD7bqK9tyS9mJimpyo8QqlNTiTpM7mBb841xtRoHpNIb7hB+vy08i631r6eCN43kOv86T1umyjb677b8lzbA4C4UfMNAPHYyRhzSpbnhltrF/sefyHpTWPMfZJ+ktRX0v6SnrTWfpBjH10ljTbGvCAXBM+Tq50+T64m+x1JstaONMY8J+kESa2NMS8rmWpwuVxaQCWWfS3RCfP0ROD8X7lUg+ck9rGdb1lrjDlVrp35OGPMo3K15dWSukg6StJVqqXDpc/ecukWvbbge0i6Kc91AaAsEHwDQDxOTNwy2UrSZN/jlyRNlAtUu0n6RdKNiVsu30t6VK5T4hFyKf9mSHpY0q3W2qW+ZU+W9Klc2r475DKRjJZ0rbV2fNp2j5cbyv5kuSYh4+UC6ZPkC74lyVr7uTFmx0TZD5d0rqRFcu2yH5Pk7/iZlTGmu1znzLcTs3pKauV7DAB1Anm+AaBM+fJ8X2+tHRhvaQAAQaDNNwAAABARgm8AAAAgIgTfAAAAQERo8w0AAABEhJpvAAAAICIE3wAAAEBECL4BAACAiBB8AwAAABEh+AYAAAAi8v+qSk+qgImDbAAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 864x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "experiment = storage.Experiment.load('../experiments/learn_firewire/firewire/2021_11_03_181436708449_c5bb2/')\n",
    "scores = experiment.scores()\n",
    "hplt.plot_training_progress(scores)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6ece2e8a-9195-43c1-bffc-f95d7d4e55b2",
   "metadata": {},
   "source": [
    "# 2. Assessing the Quality of an NN (Sect. 3 and 4 of the paper)\n",
    "\n",
    "Here we demonstrate how the modified version of the statistical model checker `modes` can be used to asses the quality of a trained agent. To this end, the nondeterminism in the model occurring during statistical model checking is resolved using the NN. We also show how the probabilistic model checker `mcsta` can be invoked on the model using Momba to check the optimal goal probability.\n",
    "\n",
    "This demonstrates the functions `check_nn` and `check_oracle` (see paper).\n",
    "\n",
    "Please note, that in the following, `modes` and `mcsta` may need several minutes to calculate the results, i.e., it may take some time until you see the result when executing the cell."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "df06a4a8-61bc-4a79-b8f4-f2f7727a3546",
   "metadata": {},
   "outputs": [],
   "source": [
    "import asyncio\n",
    "import pathlib\n",
    "\n",
    "from momba.gym import checker\n",
    "from momba.tools import modest\n",
    "\n",
    "from mogym.load_nn import load_nn\n",
    "\n",
    "# the `mcsta` probabilistic model checker of The Modest Toolset\n",
    "mcsta = modest.get_checker(accept_license=True)\n",
    "\n",
    "# the asyncio loop used internally by Jupyter\n",
    "loop = asyncio.get_event_loop()\n",
    "\n",
    "# used to load the NNs of multiple episodes\n",
    "def load_episode_nns(path):\n",
    "    import re\n",
    "    \n",
    "    policies = {}\n",
    "    for path in EXPERIMENT_PATH.glob(\"*.json\"):\n",
    "        match = re.search(\"_([0-9]+).json\", path.name)\n",
    "        if not match:\n",
    "            continue\n",
    "        episode = int(match[1])\n",
    "        policies[episode] = load_nn(path)\n",
    "    return policies"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4c345f26-fb42-4eab-9a63-3cf106d94116",
   "metadata": {},
   "source": [
    "## 2.1. Racetrack"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "6e951ff3-71e1-4ca7-ac41-abbaaf863c52",
   "metadata": {},
   "outputs": [],
   "source": [
    "from mogym import learn_racetrack\n",
    "\n",
    "EXPERIMENT_PATH = pathlib.Path(\"../experiments/learn_racetrack/racetrack2/2022_01_20_210439836184_5_fbdf8\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4b15c143",
   "metadata": {},
   "source": [
    "The optimal goal probability of `99.99%` can be checked by invoking the probabilistic model checker `mcsta` on the model. This took less than a minute on our machine."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "09b41a0f-eb79-405e-a43c-7f6855327c24",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimal Probability: 0.999955338445111\n",
      "CPU times: user 36.5 ms, sys: 13.9 ms, total: 50.5 ms\n",
      "Wall time: 6.14 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "optimal_probability = float(mcsta.check(learn_racetrack.network)[\"goalProbability\"])\n",
    "print(f\"Optimal Probability: {optimal_probability}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "19f15cbb",
   "metadata": {},
   "source": [
    "We can also check the goal reachability probability achieved by the NNs extracted after different training episodes.\n",
    "\n",
    "On our machine this took roughly two hours running four processes in parallel.\n",
    "\n",
    "Depending on your machine, you may want to run only 2 or 1 processes at the same time. The number can be configured in the following cell. Note that this does increase the time required to run the analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "07066e04-1321-4514-9f51-58976875cc9d",
   "metadata": {},
   "outputs": [],
   "source": [
    "from concurrent import futures\n",
    "\n",
    "# If you run into any issues (in particular related to memory) please try reduing the number of workers.\n",
    "pool = futures.ThreadPoolExecutor(max_workers=4)\n",
    "\n",
    "def analyze_racetrack_episodes(path):\n",
    "    options = checker.ModesOptions(additional_options=[\"-C\", \"0.99\"])\n",
    "    \n",
    "    def check_nn(item):\n",
    "        episode, nn = item\n",
    "        try:\n",
    "            result = float(checker.check_nn(learn_racetrack.network, learn_racetrack.controlled_instance, nn, options=options)[\"goalProbability\"])\n",
    "        except Exception as error:\n",
    "            print(\"Error:\", path, episode)\n",
    "            print(error)\n",
    "            raise\n",
    "        return (episode, result)\n",
    "        \n",
    "    return list(pool.map(check_nn, load_episode_nns(path).items()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "4e968ae0",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n",
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n",
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n",
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 515 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 59.7 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.9589211618257262\n",
      "  Runs used:             4820\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 515 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 262.4 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.8678233438485804\n",
      "  Runs used:             12680\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 515 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 970.6 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.7679646936656283\n",
      "  Runs used:             19260\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 158 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 201.7 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.9046343975283213\n",
      "  Runs used:             9710\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 515 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 1625.6 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.7362505978000956\n",
      "  Runs used:             20910\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 153 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 530.7 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.05519480519480519\n",
      "  Runs used:             6160\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 154 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 262.6 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.8883363471971067\n",
      "  Runs used:             11060\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 155 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 2418.5 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.589945652173913\n",
      "  Runs used:             25760\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 159 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 75.6 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.9722379603399434\n",
      "  Runs used:             3530\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 155 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 1112.5 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.6456850715746422\n",
      "  Runs used:             24450\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 154 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 1460.8 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.7610997963340123\n",
      "  Runs used:             19640\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 155 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 210.4 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.9076109936575053\n",
      "  Runs used:             9460\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 155 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 305.2 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.9228220858895706\n",
      "  Runs used:             8150\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 154 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 1274.7 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.059049079754601226\n",
      "  Runs used:             6520\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 154 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 157.6 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.9098166127292341\n",
      "  Runs used:             9270\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 154 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 2991.5 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.7498269896193772\n",
      "  Runs used:             20230\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 156 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 45.9 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.9717877094972067\n",
      "  Runs used:             3580\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n",
      "\n",
      "Peak memory usage: 154 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 1444.0 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.048555956678700364\n",
      "  Runs used:             5540\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n",
      "\n",
      "Peak memory usage: 154 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 6533.5 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.37734343837255685\n",
      "  Runs used:             25070\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n",
      "\n",
      "Peak memory usage: 153 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 5097.8 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0.25579922027290447\n",
      "  Runs used:             20520\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n",
      "[(14999, 0.7679646936656283), (16999, 0.7362505978000956), (3999, 0.9589211618257262), (6999, 0.8678233438485804), (8999, 0.589945652173913), (18999, 0.37734343837255685), (5999, 0.9046343975283213), (999, 0.05519480519480519), (17999, 0.8883363471971067), (13999, 0.7610997963340123), (4999, 0.6456850715746422), (19999, 0.9722379603399434), (11999, 0.25579922027290447), (9999, 0.9076109936575053), (15999, 0.9228220858895706), (12999, 0.7498269896193772), (2999, 0.059049079754601226), (10999, 0.9098166127292341), (1999, 0.048555956678700364), (7999, 0.9717877094972067)]\n",
      "CPU times: user 1.54 s, sys: 401 ms, total: 1.94 s\n",
      "Wall time: 2h 7min 38s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "results = analyze_racetrack_episodes(EXPERIMENT_PATH)\n",
    "print(results)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c5f8ba25",
   "metadata": {},
   "source": [
    "From this, we can extract the maximal goal reachability probability achieved by the extracted neural networks. Note that the result may not be identical to what we reported in the paper on page 10 (97.30%), because we run statistical model checking here, i.e., the results can vary within the confidence interval."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "8f63f15b-a0ea-490b-916b-362eaccf20aa",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9722379603399434"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "max([prob for _, prob in results])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8580ca96",
   "metadata": {},
   "source": [
    "The combined plot from Figure 3 b) can then be obtained by plotting the goal reachability probability calculated with `modes` for the NNs extracted after different training episodes above and the sliding mean over the training returns logged during training of the NN. Again, the goal probabilities calculated with SMC may vary a little bit."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "f150e1fd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 460.8x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "r = {}\n",
    "for episode, prob in results:\n",
    "    r[episode] = prob\n",
    "    \n",
    "results = r\n",
    "\n",
    "\n",
    "# take the grps from the analysis\n",
    "grps = []\n",
    "x_grps = []\n",
    "for episode in sorted(results):\n",
    "    x_grps.append(int(episode)+1)\n",
    "    grps.append(results[episode])\n",
    "x_grps = np.array(x_grps)/1000\n",
    "\n",
    "# load the scores from experiment\n",
    "experiment = storage.Experiment.load('../experiments/learn_racetrack/racetrack2/2022_01_20_210439836184_5_fbdf8/')\n",
    "scores = experiment.scores()\n",
    "\n",
    "# parameters for sliding mean\n",
    "max_len = len(scores)\n",
    "batch_size = 250\n",
    "# calculate the sliding mean\n",
    "d = []\n",
    "sliding_means = []\n",
    "for i, score in enumerate(scores[batch_size-1:max_len]):\n",
    "    if i < batch_size:\n",
    "        d.append(score)\n",
    "    else:\n",
    "        d[i%batch_size] = score\n",
    "    new_value = np.mean(d)\n",
    "    sliding_means.append(new_value)\n",
    "    \n",
    "# plot sliding mean and grp\n",
    "fontsize = 16\n",
    "fig, ax1 = plt.subplots(figsize=[6.4, 6])\n",
    "ax2 = ax1.twinx()\n",
    "\n",
    "x_means = (np.array(range(len(sliding_means)))+batch_size)/1000\n",
    "ax1.plot(x_means, sliding_means, label = 'racetrack', color = 'tab:blue')\n",
    "ax2.plot(x_grps, grps, linestyle = \"dotted\", color = \"red\", marker = 'o')\n",
    "\n",
    "ax1.set_xlabel('Episode [thousands]', fontsize = fontsize)\n",
    "ax1.set_ylabel('Return', fontsize = fontsize, color = 'tab:blue')\n",
    "ax1.tick_params(labelsize = fontsize)\n",
    "ax2.tick_params(labelsize=fontsize)\n",
    "ax2.set_ylabel(\"Goal Reachability Probability\", fontsize=fontsize, color = \"red\")\n",
    "\n",
    "fig.savefig(\"../racetrack-plot.pdf\", bbox_inches=\"tight\")\n",
    "fig.savefig(\"../racetrack-plot.png\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d95dd49f",
   "metadata": {
    "tags": []
   },
   "source": [
    "## 2.2. cdrive.2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "bab0ebfc-42b7-4b1a-956c-6e54908cd3df",
   "metadata": {},
   "outputs": [],
   "source": [
    "from mogym import learn_cdrive"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "eeb162f4-9e2c-4e63-bbf5-a1c22dfed929",
   "metadata": {},
   "outputs": [],
   "source": [
    "EXPERIMENT_PATH = pathlib.Path(\"../experiments/learn_cdrive/cdrive/2022_01_20_133149959763_0_664e5/\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b26f34f4",
   "metadata": {},
   "source": [
    "The optimal goal probability of `86.45%` can be checked by invoking the probabilistic model checker `mcsta` on the model. This took around half a second on our machine."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "72da3dc4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimal Probability: 0.8645657165326853\n",
      "CPU times: user 15.4 ms, sys: 16.1 ms, total: 31.6 ms\n",
      "Wall time: 592 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "optimal_probability = float(mcsta.check(learn_cdrive.network)[\"goal\"])\n",
    "print(f\"Optimal Probability: {optimal_probability}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a3bacb8-d35c-4bf1-9f37-b05b0cd5b8f3",
   "metadata": {},
   "source": [
    "We now analyze the goal reaching probability for cdrive over a subset of the training episodes, which took around one and a half minute on our machine."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "3fda0634-31d3-4b8c-939d-74aad95f5249",
   "metadata": {},
   "outputs": [],
   "source": [
    "def analyze_cdrive_episodes(path):\n",
    "    options = checker.ModesOptions(additional_options=[\"-C\", \"0.99\"])\n",
    "\n",
    "    return {\n",
    "        episode: float(checker.check_nn(learn_cdrive.network, learn_cdrive.instance, nn, options=options)[\"goal\"])\n",
    "        for episode, nn in load_episode_nns(path).items()\n",
    "    }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "43036142-f7a8-4562-afd4-455c5eac0bee",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: A Momba Model is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 179 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 23.9 s\n",
      "\n",
      "+ Property goal\n",
      "  Estimated probability: 0.8075739644970414\n",
      "  Runs used:             16900\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: A Momba Model is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 175 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 10.5 s\n",
      "\n",
      "+ Property goal\n",
      "  Estimated probability: 0.8661202185792349\n",
      "  Runs used:             12810\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: A Momba Model is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 175 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 5.0 s\n",
      "\n",
      "+ Property goal\n",
      "  Estimated probability: 0.8652680652680653\n",
      "  Runs used:             12870\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: A Momba Model is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 174 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 4.7 s\n",
      "\n",
      "+ Property goal\n",
      "  Estimated probability: 0.865990639625585\n",
      "  Runs used:             12820\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: A Momba Model is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 175 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 4.6 s\n",
      "\n",
      "+ Property goal\n",
      "  Estimated probability: 0.8684877276326207\n",
      "  Runs used:             12630\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: A Momba Model is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 174 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 5.0 s\n",
      "\n",
      "+ Property goal\n",
      "  Estimated probability: 0.8611532625189682\n",
      "  Runs used:             13180\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: A Momba Model is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 174 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 5.1 s\n",
      "\n",
      "+ Property goal\n",
      "  Estimated probability: 0.8603018867924528\n",
      "  Runs used:             13250\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: A Momba Model is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 174 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 4.9 s\n",
      "\n",
      "+ Property goal\n",
      "  Estimated probability: 0.8666405638214565\n",
      "  Runs used:             12770\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: A Momba Model is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 173 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 4.8 s\n",
      "\n",
      "+ Property goal\n",
      "  Estimated probability: 0.8645510835913313\n",
      "  Runs used:             12920\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "model.jani:model: info: A Momba Model is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "\n",
      "Peak memory usage: 175 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 5.1 s\n",
      "\n",
      "+ Property goal\n",
      "  Estimated probability: 0.8618431073876618\n",
      "  Runs used:             13130\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.010000000000000009\n",
      "\n",
      "CPU times: user 406 ms, sys: 123 ms, total: 529 ms\n",
      "Wall time: 1min 25s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "results_cdrive = analyze_cdrive_episodes(EXPERIMENT_PATH)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "e149e4d3-f114-4aa7-9223-83768cc3e544",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{999: 0.8075739644970414,\n",
       " 1999: 0.8661202185792349,\n",
       " 6999: 0.8652680652680653,\n",
       " 2999: 0.865990639625585,\n",
       " 8999: 0.8684877276326207,\n",
       " 7999: 0.8611532625189682,\n",
       " 9999: 0.8603018867924528,\n",
       " 5999: 0.8666405638214565,\n",
       " 4999: 0.8645510835913313,\n",
       " 3999: 0.8618431073876618}"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_cdrive"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c449a1ca",
   "metadata": {},
   "source": [
    "From that the maximal goal reachability probability achieved by the extracted neural networks can be obtained. Again, the result may differ from what we reported on page 9 at the bottom (86.57%) because we perform statistical model checking and the results may vary within the confidence interval."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "b03a209c-a5e7-4ba9-98e2-61fb8a0ea114",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8684877276326207"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "max(results_cdrive.values())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5e2902d7",
   "metadata": {},
   "source": [
    "The combined plot in Figure 3 a) can then again be produced from the intermediate results as above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "4d66a724-aca1-4536-a729-61bc68c8de4d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 460.8x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# take the grps from the analysis\n",
    "results = results_cdrive\n",
    "grps = []\n",
    "x_grps = []\n",
    "for episode in sorted(results_cdrive):\n",
    "    x_grps.append(episode+1)\n",
    "    grps.append(results[episode])\n",
    "x_grps = np.array(x_grps)/1000\n",
    "\n",
    "# load the scores from experiment\n",
    "experiment = storage.Experiment.load('../experiments/learn_cdrive/cdrive/2022_01_20_133149959763_0_664e5/')\n",
    "scores = experiment.scores()\n",
    "\n",
    "# parameters for sliding mean\n",
    "max_len = len(scores)\n",
    "batch_size = 250\n",
    "# calculate the sliding mean\n",
    "d = []\n",
    "sliding_means = []\n",
    "for i, score in enumerate(scores[batch_size-1:max_len]):\n",
    "    if i < batch_size:\n",
    "        d.append(score)\n",
    "    else:\n",
    "        d[i%batch_size] = score\n",
    "    new_value = np.mean(d)\n",
    "    sliding_means.append(new_value)\n",
    "    \n",
    "# plot sliding mean and grp\n",
    "fontsize = 16\n",
    "fig, ax1 = plt.subplots(figsize=[6.4, 6])\n",
    "ax2 = ax1.twinx()\n",
    "\n",
    "x_means = (np.array(range(len(sliding_means)))+batch_size)/1000\n",
    "ax1.plot(x_means, sliding_means, label = 'cdrive.2', color = 'tab:blue')\n",
    "ax2.plot(x_grps, grps, linestyle = \"dotted\", color = \"red\", marker = 'o')\n",
    "\n",
    "ax1.set_xlabel('Episode [thousands]', fontsize = fontsize)\n",
    "ax1.set_ylabel('Return', fontsize = fontsize, color = 'tab:blue')\n",
    "ax1.tick_params(labelsize = fontsize)\n",
    "ax2.tick_params(labelsize=fontsize)\n",
    "ax2.set_ylabel(\"Goal Reachability Probability\", fontsize=fontsize, color = \"red\")\n",
    "\n",
    "fig.savefig(\"../cdrive-plot.pdf\", bbox_inches=\"tight\")\n",
    "fig.savefig(\"../cdrive-plot.png\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b445e765",
   "metadata": {},
   "source": [
    "# 3. Assessing the Quality of an Arbitrary Decision Agent"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "abbbef63",
   "metadata": {},
   "source": [
    "In addition to checking the quality of an NN, MoGym also makes it possible to check arbitrary decision agents given as a Python function by connecting it to `modes` to resolve nondeterminism. For demonstration purposes, we here use a very simple decision agent, which always chooses the first applicable action. Clearly, this results in a very bad goal probability of 0. The execution took about 1 minute on our machine, please wait for the output even if the notebook already pretends to be idle."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "31e27e03-177d-41c1-b856-dc5bf49d1c1b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Task pending name='Task-4' coro=<check_oracle_async() running at /home/vscode/.local/lib/python3.9/site-packages/momba/gym/checker.py:174>>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Address: localhost:42057\n",
      "\n",
      "\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The Modest Toolset (www.modestchecker.net), version v3.1.107-g50c678139.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "model.jani:model: info: Featured Racetrack is an MDP model and will be simulated as an MDP.\n",
      "model.jani: info: Using default value of 0.95 for the confidence parameter.\n",
      "model.jani: info: Using default value of 0.01 for the absolute half-width parameter.\n",
      "New connection from ('127.0.0.1', 42998).\n",
      "Features: 15, Actions: 9\n",
      "\n",
      "Peak memory usage: 82 MB\n",
      "Analysis results for model.jani\n",
      "Status:          Finished\n",
      "Simulation time: 37.3 s\n",
      "\n",
      "+ Property goalProbability\n",
      "  Estimated probability: 0\n",
      "  Runs used:             490\n",
      "  Run type:              MDP\n",
      "  Status:                Finished\n",
      "\n",
      "  + Error bounds\n",
      "    Statement: Adaptive: P(error > ε) < δ\n",
      "    ε:         0.01\n",
      "    δ:         0.050000000000000044\n",
      "\n"
     ]
    }
   ],
   "source": [
    "def oracle(state, available):\n",
    "    \"\"\" This is of course a very dumb driver! \"\"\"\n",
    "    return 1\n",
    "\n",
    "# We use the asynchronous API here for compatibility with Jupyter Lab. As a result, it\n",
    "# may seem as if the cell has terminated although this is not the case. You will see the\n",
    "# output of `modes` which indicates the probability.\n",
    "loop.create_task(checker.check_oracle_async(learn_racetrack.network, learn_racetrack.controlled_instance, oracle))"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "71cf75d41eec4e5aac84aa4900c2fceef1cea5e3a49fd988c576cc53e7371299"
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}