rlberry_evaluate_and_optimize_agent.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "rlberry_evaluate_and_optimize_agent.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3.7.11 64-bit ('rlberry': conda)"
    },
    "language_info": {
      "name": "python",
      "version": "3.7.11",
      "mimetype": "text/x-python",
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "pygments_lexer": "ipython3",
      "nbconvert_exporter": "python",
      "file_extension": ".py"
    },
    "interpreter": {
      "hash": "77626d6da4f24721c6cde6592fa6521e12f75aa06b21b1ac25df0859aa7f2530"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/rlberry-py/notebooks/blob/main/rlberry_evaluate_and_optimize_agent.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Evaluate an agent and optimize its hyperparameters with `rlberry`"
      ],
      "metadata": {
        "id": "sjoqYioOnAQB"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Colab Setup"
      ],
      "metadata": {
        "id": "MsUygvTOnJO_"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "source": [
        "from IPython import get_ipython\n",
        "COLAB = False\n",
        "if 'google.colab' in str(get_ipython()):\n",
        "    COLAB = True\n",
        "\n",
        "if COLAB:\n",
        "    # install rlberry library\n",
        "    !pip install git+https://github.com/rlberry-py/rlberry.git@v0.2.1#egg=rlberry[default] > /dev/null 2>&1\n",
        "\n",
        "    !apt-get install -y xvfb python-opengl ffmpeg > /dev/null 2>&1\n",
        "\n",
        "    print(\"\")\n",
        "    print(\" ~~~  Libraries installed, please restart the runtime! ~~~ \")\n",
        "    print(\"\")\n"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            " ~~~  Libraries installed, please restart the runtime! ~~~ \n",
            "\n"
          ]
        }
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mh5ZCgwlmxJ1",
        "outputId": "0f907839-fcc4-4b8a-a764-7f653d953b0d"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## `AgentManager`: training several instances of an agent\n",
        "\n",
        "The class `AgentManager` can be used to train several instances of an agent (in parallel), optimize its hyperparameters and analyze its performance.\n",
        "\n",
        "Check the code below to learn how!"
      ],
      "metadata": {
        "id": "kCaQFlRZnLZH"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "source": [
        "from rlberry.envs import gym_make\n",
        "from rlberry.agents.torch.reinforce import REINFORCEAgent\n",
        "from rlberry.manager import AgentManager, plot_writer_data\n",
        "\n",
        "\n",
        "# Environment\n",
        "env_ctor = gym_make\n",
        "env_kwargs = dict(id='CartPole-v0')\n",
        "env = env_ctor(**env_kwargs)\n",
        "\n",
        "# Initial set of parameters\n",
        "params = dict(\n",
        "    gamma=0.99,\n",
        "    horizon=200\n",
        ")\n",
        "\n",
        "# Create AgentManager to fit 2 instances of REINFORCE in parallel.\n",
        "stats = AgentManager(REINFORCEAgent,\n",
        "                   (env_ctor, env_kwargs),\n",
        "                   fit_budget=500,   # number of episodes\n",
        "                   init_kwargs=params,\n",
        "                   eval_kwargs=dict(eval_horizon=200),\n",
        "                   n_fit=2,\n",
        "                   parallelization='thread')\n",
        "\n",
        "# Fit the 4 instances\n",
        "stats.fit()\n",
        "\n",
        "# The fit() method of REINFORCEAgent writes data to a DefaultWriter (rlberry.utils.writer)\n",
        "# object, which can be plotted with the method below.\n",
        "plot_writer_data(stats, tag='episode_rewards')"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] OpenGL_accelerate module loaded \n",
            "[INFO] Using accelerated ArrayDatatype \n",
            "[INFO] Generating grammar tables from /usr/lib/python3.7/lib2to3/Grammar.txt \n",
            "[INFO] Generating grammar tables from /usr/lib/python3.7/lib2to3/PatternGrammar.txt \n",
            "[INFO] Running AgentManager fit() for REINFORCE...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/rlberry/agents/torch/reinforce/reinforce.py:188: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at  ../torch/csrc/utils/tensor_new.cpp:201.)\n",
            "  states = torch.FloatTensor(self.memory.states).to(self.device)\n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 120 | episode_rewards = 34.0 | dw_time_elapsed = 3.0056547020000153 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 129 | episode_rewards = 19.0 | dw_time_elapsed = 3.0063819619999776 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 267 | episode_rewards = 17.0 | dw_time_elapsed = 6.027439900000019 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 277 | episode_rewards = 21.0 | dw_time_elapsed = 6.036842956999976 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 385 | episode_rewards = 20.0 | dw_time_elapsed = 9.051516378000002 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 386 | episode_rewards = 57.0 | dw_time_elapsed = 9.089783515999983 |  \n",
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-72b79457-3734-42a2-bcba-8b3d516bfe89\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>name</th>\n",
              "      <th>tag</th>\n",
              "      <th>value</th>\n",
              "      <th>global_step</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>11.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>75.0</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>24.0</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>19.0</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>14.0</td>\n",
              "      <td>5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>995</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>51.0</td>\n",
              "      <td>496</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>996</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>14.0</td>\n",
              "      <td>497</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>997</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>12.0</td>\n",
              "      <td>498</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>998</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>17.0</td>\n",
              "      <td>499</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>999</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>12.0</td>\n",
              "      <td>500</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>1000 rows × 4 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-72b79457-3734-42a2-bcba-8b3d516bfe89')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-72b79457-3734-42a2-bcba-8b3d516bfe89 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-72b79457-3734-42a2-bcba-8b3d516bfe89');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "          name              tag  value global_step\n",
              "0    REINFORCE  episode_rewards   11.0           1\n",
              "1    REINFORCE  episode_rewards   75.0           2\n",
              "2    REINFORCE  episode_rewards   24.0           3\n",
              "3    REINFORCE  episode_rewards   19.0           4\n",
              "4    REINFORCE  episode_rewards   14.0           5\n",
              "..         ...              ...    ...         ...\n",
              "995  REINFORCE  episode_rewards   51.0         496\n",
              "996  REINFORCE  episode_rewards   14.0         497\n",
              "997  REINFORCE  episode_rewards   12.0         498\n",
              "998  REINFORCE  episode_rewards   17.0         499\n",
              "999  REINFORCE  episode_rewards   12.0         500\n",
              "\n",
              "[1000 rows x 4 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 2
        }
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "4f9QzuOQniNY",
        "outputId": "810c354a-a5fc-44d9-c21a-9ccac3861ac2"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## `AgentManager`: optimizing hyperparameters\n",
        "\n",
        "The agent was not able to learn with the previous set of hyperparameters. Let's see if we can make it work by optimizing these parameters!\n",
        "\n",
        "`REINFORCEAgent` implements a method `sample_hyperparameters`, which allows `AgentManager` to search for the best configuration of hyperparameters. In the class `REINFORCEAgent`, we have:\n",
        "\n",
        "```python\n",
        "@classmethod\n",
        "def sample_parameters(cls, trial):\n",
        "    \"\"\"\n",
        "    Sample hyperparameters for hyperparam optimization using\n",
        "    Optuna (https://optuna.org/)\n",
        "\n",
        "    Note: only the kwargs sent to __init__ are optimized. Make sure to\n",
        "    include in the Agent constructor all \"optimizable\" parameters.\n",
        "\n",
        "    Parameters\n",
        "    ----------\n",
        "    trial: optuna.trial\n",
        "    \"\"\"\n",
        "    batch_size = trial.suggest_categorical('batch_size', [1, 4, 8, 16, 32])\n",
        "    gamma = trial.suggest_categorical('gamma', [0.9, 0.95, 0.99])\n",
        "    learning_rate = trial.suggest_loguniform('learning_rate', 1e-5, 1)\n",
        "    entr_coef = trial.suggest_loguniform('entr_coef', 1e-8, 0.1)\n",
        "\n",
        "    return {\n",
        "            'batch_size': batch_size,\n",
        "            'gamma': gamma,\n",
        "            'learning_rate': learning_rate,\n",
        "            'entr_coef': entr_coef,\n",
        "            }\n",
        "```\n"
      ],
      "metadata": {
        "id": "DjDWGNRxpF4z"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "source": [
        "# Run optimization and print results\n",
        "stats.optimize_hyperparams(\n",
        "    n_trials=100,\n",
        "    timeout=120,   # stop after 120 seconds\n",
        "    n_fit=1,\n",
        "    sampler_method='optuna_default'\n",
        ")\n",
        "\n",
        "print(stats.best_hyperparams)\n",
        "\n",
        "# Calling fit() again will train the agent with the optimized parameters\n",
        "stats.fit()\n",
        "plot_writer_data(stats, tag='episode_rewards')"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Connected to rlberry_data/temp/manager_data/REINFORCE_09504970db34c1faa05d879251f00b2d/optuna_data.db (sqlite3 version = 2.6.0)\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\u001b[32m[I 2022-02-20 21:29:58,714]\u001b[0m A new study created in RDB with name: no-name-a402e3bb-938d-4fd1-926e-ea101e623eec\u001b[0m\n",
            "[INFO] Saved AgentManager(REINFORCE) using pickle. \n",
            "[INFO] Running AgentManager fit() for optim...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n",
            "[INFO] [eval]... simulation 1/2 \n",
            "[INFO] [eval]... simulation 2/2 \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\u001b[32m[I 2022-02-20 21:30:04,224]\u001b[0m Trial 0 finished with value: 26.75 and parameters: {'batch_size': 32, 'gamma': 0.9, 'learning_rate': 0.00040262365985170394, 'entr_coef': 2.6012852598162817e-05}. Best is trial 0 with value: 26.75.\u001b[0m\n",
            "[INFO] Running AgentManager fit() for optim...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n",
            "[INFO] [eval]... simulation 1/2 \n",
            "[INFO] [eval]... simulation 2/2 \n",
            "\u001b[32m[I 2022-02-20 21:30:08,050]\u001b[0m Trial 1 finished with value: 23.25 and parameters: {'batch_size': 8, 'gamma': 0.99, 'learning_rate': 5.270413158545329e-05, 'entr_coef': 0.003519307834216726}. Best is trial 0 with value: 26.75.\u001b[0m\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] Running AgentManager fit() for optim...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] [eval]... simulation 1/2 \n",
            "[INFO] [eval]... simulation 2/2 \n",
            "\u001b[32m[I 2022-02-20 21:30:29,690]\u001b[0m Trial 2 finished with value: 140.95 and parameters: {'batch_size': 8, 'gamma': 0.9, 'learning_rate': 0.005575408175424816, 'entr_coef': 0.013755922923785425}. Best is trial 2 with value: 140.95.\u001b[0m\n",
            "[INFO] Running AgentManager fit() for optim...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n",
            "[INFO] [eval]... simulation 1/2 \n",
            "[INFO] [eval]... simulation 2/2 \n",
            "\u001b[32m[I 2022-02-20 21:30:33,229]\u001b[0m Trial 3 finished with value: 23.55 and parameters: {'batch_size': 4, 'gamma': 0.99, 'learning_rate': 6.195313593200499e-05, 'entr_coef': 1.447347136328782e-07}. Best is trial 2 with value: 140.95.\u001b[0m\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] Running AgentManager fit() for optim...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n",
            "[INFO] [eval]... simulation 1/2 \n",
            "[INFO] [eval]... simulation 2/2 \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\u001b[32m[I 2022-02-20 21:30:37,232]\u001b[0m Trial 4 finished with value: 26.75 and parameters: {'batch_size': 4, 'gamma': 0.95, 'learning_rate': 6.261943299509413e-05, 'entr_coef': 0.0017773139011075033}. Best is trial 2 with value: 140.95.\u001b[0m\n",
            "[INFO] Running AgentManager fit() for optim...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n",
            "[INFO] [eval]... simulation 1/2 \n",
            "[INFO] [eval]... simulation 2/2 \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\u001b[32m[I 2022-02-20 21:30:41,281]\u001b[0m Trial 5 finished with value: 28.950000000000003 and parameters: {'batch_size': 4, 'gamma': 0.99, 'learning_rate': 6.685205273650854e-05, 'entr_coef': 4.9595184594866164e-08}. Best is trial 2 with value: 140.95.\u001b[0m\n",
            "[INFO] Running AgentManager fit() for optim...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] [eval]... simulation 1/2 \n",
            "[INFO] [eval]... simulation 2/2 \n",
            "\u001b[32m[I 2022-02-20 21:30:56,877]\u001b[0m Trial 6 finished with value: 132.8 and parameters: {'batch_size': 16, 'gamma': 0.95, 'learning_rate': 0.008794506033769557, 'entr_coef': 0.005350920318904114}. Best is trial 2 with value: 140.95.\u001b[0m\n",
            "[INFO] Running AgentManager fit() for optim...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] [eval]... simulation 1/2 \n",
            "[INFO] [eval]... simulation 2/2 \n",
            "\u001b[32m[I 2022-02-20 21:31:10,149]\u001b[0m Trial 7 finished with value: 159.0 and parameters: {'batch_size': 16, 'gamma': 0.99, 'learning_rate': 0.006668270459248973, 'entr_coef': 0.004187487444773686}. Best is trial 7 with value: 159.0.\u001b[0m\n",
            "[INFO] Running AgentManager fit() for optim...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] [eval]... simulation 1/2 \n",
            "[INFO] [eval]... simulation 2/2 \n",
            "\u001b[32m[I 2022-02-20 21:31:18,089]\u001b[0m Trial 8 finished with value: 102.4 and parameters: {'batch_size': 32, 'gamma': 0.95, 'learning_rate': 0.006617169979479383, 'entr_coef': 7.672114252191604e-07}. Best is trial 7 with value: 159.0.\u001b[0m\n",
            "[INFO] Running AgentManager fit() for optim...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n",
            "[INFO] [eval]... simulation 1/2 \n",
            "[INFO] [eval]... simulation 2/2 \n",
            "\u001b[32m[I 2022-02-20 21:31:20,554]\u001b[0m Trial 9 finished with value: 9.3 and parameters: {'batch_size': 1, 'gamma': 0.95, 'learning_rate': 0.01917447400577418, 'entr_coef': 0.0012451299911390079}. Best is trial 7 with value: 159.0.\u001b[0m\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] Running AgentManager fit() for optim...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n",
            "[INFO] [eval]... simulation 1/2 \n",
            "[INFO] [eval]... simulation 2/2 \n",
            "\u001b[32m[I 2022-02-20 21:31:22,464]\u001b[0m Trial 10 finished with value: 9.3 and parameters: {'batch_size': 16, 'gamma': 0.99, 'learning_rate': 0.9222046079249648, 'entr_coef': 4.16636384781376e-05}. Best is trial 7 with value: 159.0.\u001b[0m\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] Running AgentManager fit() for optim...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n",
            "[INFO] [eval]... simulation 1/2 \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] [eval]... simulation 2/2 \n",
            "\u001b[32m[I 2022-02-20 21:31:31,162]\u001b[0m Trial 11 finished with value: 45.15 and parameters: {'batch_size': 8, 'gamma': 0.9, 'learning_rate': 0.12023561752427234, 'entr_coef': 0.0982326618644128}. Best is trial 7 with value: 159.0.\u001b[0m\n",
            "[INFO] Running AgentManager fit() for optim...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] [eval]... simulation 1/2 \n",
            "[INFO] [eval]... simulation 2/2 \n",
            "\u001b[32m[I 2022-02-20 21:31:38,602]\u001b[0m Trial 12 finished with value: 94.4 and parameters: {'batch_size': 8, 'gamma': 0.9, 'learning_rate': 0.0010369500912535325, 'entr_coef': 0.06768982169343572}. Best is trial 7 with value: 159.0.\u001b[0m\n",
            "[INFO] Running AgentManager fit() for optim...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] [eval]... simulation 1/2 \n",
            "[INFO] [eval]... simulation 2/2 \n",
            "\u001b[32m[I 2022-02-20 21:31:59,510]\u001b[0m Trial 13 finished with value: 200.0 and parameters: {'batch_size': 16, 'gamma': 0.9, 'learning_rate': 0.04494654060829228, 'entr_coef': 0.00011618534375752914}. Best is trial 13 with value: 200.0.\u001b[0m\n",
            "[INFO] Number of finished trials: 14 \n",
            "[INFO] Best trial: \n",
            "[INFO] Value: 200.0 \n",
            "[INFO] Params: \n",
            "[INFO]     batch_size: 16 \n",
            "[INFO]     entr_coef: 0.00011618534375752914 \n",
            "[INFO]     gamma: 0.9 \n",
            "[INFO]     learning_rate: 0.04494654060829228 \n",
            "[INFO] Running AgentManager fit() for REINFORCE...  \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "{'batch_size': 16, 'entr_coef': 0.00011618534375752914, 'gamma': 0.9, 'learning_rate': 0.04494654060829228}\n",
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 82 | episode_rewards = 64.0 | dw_time_elapsed = 3.0380977460000054 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 81 | episode_rewards = 96.0 | dw_time_elapsed = 3.0292299780000462 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 111 | episode_rewards = 130.0 | dw_time_elapsed = 6.070405113999982 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 145 | episode_rewards = 142.0 | dw_time_elapsed = 6.1262245700000335 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 134 | episode_rewards = 137.0 | dw_time_elapsed = 9.179841453999984 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 164 | episode_rewards = 200.0 | dw_time_elapsed = 9.186684895000042 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 182 | episode_rewards = 135.0 | dw_time_elapsed = 12.236836961000051 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 155 | episode_rewards = 200.0 | dw_time_elapsed = 12.270595807000007 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 205 | episode_rewards = 154.0 | dw_time_elapsed = 15.308453755000016 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 174 | episode_rewards = 200.0 | dw_time_elapsed = 15.436554344000001 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 230 | episode_rewards = 152.0 | dw_time_elapsed = 18.315290884000035 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 192 | episode_rewards = 200.0 | dw_time_elapsed = 18.60918541000001 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 257 | episode_rewards = 121.0 | dw_time_elapsed = 21.435425484000007 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 209 | episode_rewards = 200.0 | dw_time_elapsed = 21.646667006999962 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 285 | episode_rewards = 110.0 | dw_time_elapsed = 24.468236302000037 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 226 | episode_rewards = 200.0 | dw_time_elapsed = 24.674789983999972 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 314 | episode_rewards = 117.0 | dw_time_elapsed = 27.489112360000036 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 243 | episode_rewards = 200.0 | dw_time_elapsed = 27.69208366199996 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 343 | episode_rewards = 116.0 | dw_time_elapsed = 30.504377558000044 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 260 | episode_rewards = 200.0 | dw_time_elapsed = 30.792127753999978 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 373 | episode_rewards = 100.0 | dw_time_elapsed = 33.512755008 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 277 | episode_rewards = 200.0 | dw_time_elapsed = 33.948474490999956 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 412 | episode_rewards = 83.0 | dw_time_elapsed = 36.53268694600001 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 294 | episode_rewards = 200.0 | dw_time_elapsed = 37.09672404899999 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 450 | episode_rewards = 112.0 | dw_time_elapsed = 39.56612000300004 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 311 | episode_rewards = 200.0 | dw_time_elapsed = 40.19385183099996 |  \n",
            "[INFO] [REINFORCE[worker: 1]] | max_global_step = 481 | episode_rewards = 125.0 | dw_time_elapsed = 42.59069258100004 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 328 | episode_rewards = 200.0 | dw_time_elapsed = 43.288428883999984 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 363 | episode_rewards = 200.0 | dw_time_elapsed = 46.319542361 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 413 | episode_rewards = 200.0 | dw_time_elapsed = 49.34864978899998 |  \n",
            "[INFO] [REINFORCE[worker: 0]] | max_global_step = 469 | episode_rewards = 200.0 | dw_time_elapsed = 52.37240330599997 |  \n",
            "[INFO] ... trained! \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n",
            "[INFO] Could not find least used device (nvidia-smi might be missing), use cuda:0 instead \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n",
            "INFO: Making new env: CartPole-v0\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-430ad4e5-a2cf-49bf-a7f2-6fad230cc60a\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>name</th>\n",
              "      <th>tag</th>\n",
              "      <th>value</th>\n",
              "      <th>global_step</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>15.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>39.0</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>17.0</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>13.0</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>10.0</td>\n",
              "      <td>5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>995</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>128.0</td>\n",
              "      <td>496</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>996</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>136.0</td>\n",
              "      <td>497</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>997</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>120.0</td>\n",
              "      <td>498</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>998</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>135.0</td>\n",
              "      <td>499</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>999</th>\n",
              "      <td>REINFORCE</td>\n",
              "      <td>episode_rewards</td>\n",
              "      <td>138.0</td>\n",
              "      <td>500</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>1000 rows × 4 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-430ad4e5-a2cf-49bf-a7f2-6fad230cc60a')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-430ad4e5-a2cf-49bf-a7f2-6fad230cc60a button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-430ad4e5-a2cf-49bf-a7f2-6fad230cc60a');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "          name              tag  value global_step\n",
              "0    REINFORCE  episode_rewards   15.0           1\n",
              "1    REINFORCE  episode_rewards   39.0           2\n",
              "2    REINFORCE  episode_rewards   17.0           3\n",
              "3    REINFORCE  episode_rewards   13.0           4\n",
              "4    REINFORCE  episode_rewards   10.0           5\n",
              "..         ...              ...    ...         ...\n",
              "995  REINFORCE  episode_rewards  128.0         496\n",
              "996  REINFORCE  episode_rewards  136.0         497\n",
              "997  REINFORCE  episode_rewards  120.0         498\n",
              "998  REINFORCE  episode_rewards  135.0         499\n",
              "999  REINFORCE  episode_rewards  138.0         500\n",
              "\n",
              "[1000 rows x 4 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "-HttJnDVpk-5",
        "outputId": "afebb6c1-f147-40d6-d052-ab5067cc0616"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "source": [
        "# Visualize the agent\n",
        "\n",
        "import rlberry.colab_utils.display_setup\n",
        "from rlberry.colab_utils.display_setup import show_video\n",
        "from gym.wrappers import Monitor\n",
        "\n",
        "# wrap env\n",
        "env = Monitor(env, \"./gym-results\", force=True, video_callable=lambda episode: True)\n",
        "\n",
        "# take one of the trained agents\n",
        "agent = stats.agent_handlers[0]\n",
        "\n",
        "for episode in range(1):\n",
        "    done = False\n",
        "    state = env.reset()\n",
        "    while not done:\n",
        "        action = agent.policy(state)\n",
        "        state, reward, done, info = env.step(action)\n",
        "env.close()\n",
        "show_video(directory=\"./gym-results\")"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO: Clearing 4 monitor files from previous run (because force=True was provided)\n",
            "INFO: Starting new video recorder writing to /content/gym-results/openaigym.video.1.76.video000000.mp4\n",
            "INFO: Finished writing results. You can upload them to the scoreboard via gym.upload('/content/gym-results')\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<video alt=\"gym-results/openaigym.video.1.76.video000000.mp4\" autoplay\n",
              "                      loop controls style=\"height: 400px;\">\n",
              "                    <source src=\"data:video/mp4;base64,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\" type=\"video/mp4\" />\n",
              "                 </video>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {}
        }
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 474
        },
        "id": "e10uhaCpwZ2L",
        "outputId": "eeba3686-ccb6-4b8d-d6c1-81c4cd0c1876"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "source": [
        "# clear data\n",
        "stats.clear_output_dir()"
      ],
      "outputs": [],
      "metadata": {
        "id": "babnUP4bP8jZ"
      }
    }
  ]
}