LIB_Discrete.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "qF21cOz17bPL"
      },
      "outputs": [],
      "source": [
        "from IPython.display import clear_output\n",
        "\n",
        "!pip install -U \"ray[rllib]\"\n",
        "!pip install tianshou\n",
        "!pip install \"stable-baselines3[extra]>=2.0.0a4\"\n",
        "!pip install mat4py\n",
        "\n",
        "clear_output()"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "V3XZgWVE5nE5"
      },
      "source": [
        "# LIBPackEnv"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 322
        },
        "id": "aw0t5NkH5m4k",
        "outputId": "b098db97-c213-428f-c99f-b281012a972b"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "d:\\Users\\amirh\\anaconda3\\lib\\site-packages\\gymnasium\\spaces\\box.py:130: UserWarning: \u001b[33mWARN: Box bound precision lowered by casting to float32\u001b[0m\n",
            "  gym.logger.warn(f\"Box bound precision lowered by casting to {self.dtype}\")\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1800x400 with 3 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "import cv2\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import matplotlib.patches as patches\n",
        "from LIBEnv.LIB_gym import LIBPackEnv\n",
        "import seaborn as sns\n",
        "sns.set_theme()\n",
        "\n",
        "env = LIBPackEnv('./LIBEnv/PANmodel.mat', number_of_cells=6, T=25, dt=1, use_switch=True)\n",
        "\n",
        "pack_res=[]\n",
        "rewards = []\n",
        "pack_res.append(env.reset(0)[0].reshape(3,6))\n",
        "done = False\n",
        "while not done:\n",
        "    next_state, reward, done, _, info = env.step(np.ones(7))\n",
        "    pack_res.append(next_state.reshape(3,6))\n",
        "    rewards.append(reward)\n",
        "\n",
        "pack_res = np.array(pack_res)\n",
        "plt.figure(figsize=(18,4))\n",
        "plt.subplot(1,3,1)\n",
        "plt.plot(pack_res[:,0,:])\n",
        "plt.xlabel('Time (s)')\n",
        "plt.ylabel('Cell SoCs (%)')\n",
        "plt.grid('on')\n",
        "plt.subplot(1,3,2)\n",
        "plt.plot(pack_res[:,1,:]*(env.max_voltage-env.min_voltage)+env.min_voltage)\n",
        "plt.xlabel('Time (s)')\n",
        "plt.ylabel('Cell Voltages (V)')\n",
        "plt.grid('on')\n",
        "plt.subplot(1,3,3)\n",
        "plt.plot(pack_res[:,2,:]*env.T)\n",
        "plt.xlabel('Time (s)')\n",
        "plt.ylabel('Cell Core Temps (°C)')\n",
        "plt.grid('on')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/plain": [
              "array([0, 0, 0, 1, 0, 0, 1], dtype=int64)"
            ]
          },
          "execution_count": 4,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "env.action_space.sample()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Tensorboard "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "#@title Tensorboard - click the refresh button once training is running\n",
        "\n",
        "%load_ext tensorboard\n",
        "%tensorboard --logdir logs"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "0DcQa_RX5lF1"
      },
      "source": [
        "# Discrete stable baseline3"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {},
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "d:\\Users\\amirh\\anaconda3\\lib\\site-packages\\torch\\utils\\tensorboard\\__init__.py:4: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.\n",
            "  if not hasattr(tensorboard, \"__version__\") or LooseVersion(\n"
          ]
        }
      ],
      "source": [
        "from stable_baselines3.common.callbacks import BaseCallback\n",
        "\n",
        "class SaveOnBestTrainingRewardCallback(BaseCallback):\n",
        "    \"\"\"\n",
        "    Callback for saving a model (the check is done every ``check_freq`` steps)\n",
        "    based on the training reward (in practice, we recommend using ``EvalCallback``).\n",
        "\n",
        "    :param check_freq:\n",
        "    :param log_dir: Path to the folder where the model will be saved.\n",
        "      It must contains the file created by the ``Monitor`` wrapper.\n",
        "    :param verbose: Verbosity level: 0 for no output, 1 for info messages, 2 for debug messages\n",
        "    \"\"\"\n",
        "    def __init__(self, check_freq: int, log_dir: str, verbose: int = 1):\n",
        "        super(SaveOnBestTrainingRewardCallback, self).__init__(verbose)\n",
        "        self.check_freq = check_freq\n",
        "        self.log_dir = log_dir\n",
        "        self.save_path = os.path.join(log_dir, \"best_model\")\n",
        "        self.best_mean_reward = -np.inf\n",
        "\n",
        "    def _init_callback(self) -> None:\n",
        "       pass\n",
        "        # Create folder if needed\n",
        "        # if self.save_path is not None:\n",
        "        #     os.makedirs(self.save_path, exist_ok=True)\n",
        "\n",
        "    def _on_step(self) -> bool:\n",
        "        if self.n_calls % self.check_freq == 0:\n",
        "\n",
        "          # Retrieve training reward\n",
        "          x, y = ts2xy(load_results(self.log_dir), \"timesteps\")\n",
        "          if len(x) > 0:\n",
        "              # Mean training reward over the last 100 episodes\n",
        "              mean_reward = np.mean(y[-100:])\n",
        "              if self.verbose >= 1:\n",
        "                print(f\"Num timesteps: {self.num_timesteps}\")\n",
        "                print(f\"Best mean reward: {self.best_mean_reward:.2f} - Last mean reward per episode: {mean_reward:.2f}\")\n",
        "\n",
        "              # New best model, you could save the agent here\n",
        "              if mean_reward > self.best_mean_reward:\n",
        "                  self.best_mean_reward = mean_reward\n",
        "                  # Example for saving best model\n",
        "                  if self.verbose >= 1:\n",
        "                    print(f\"Saving new best model to {self.save_path}\")\n",
        "                  self.model.save(self.save_path)\n",
        "\n",
        "        return True"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "VgCQlkUqtOa-",
        "outputId": "36705df8-d277-4599-e999-994af5dfeac6"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Using cuda device\n",
            "Wrapping the env in a DummyVecEnv.\n",
            "Logging to logs\\PPO_7\n",
            "----------------------------------\n",
            "| rollout/           |           |\n",
            "|    ep_len_mean     | 1.16e+03  |\n",
            "|    ep_rew_mean     | -4.02e+04 |\n",
            "| time/              |           |\n",
            "|    fps             | 91        |\n",
            "|    iterations      | 1         |\n",
            "|    time_elapsed    | 22        |\n",
            "|    total_timesteps | 2048      |\n",
            "----------------------------------\n",
            "------------------------------------------\n",
            "| rollout/                |              |\n",
            "|    ep_len_mean          | 1.06e+03     |\n",
            "|    ep_rew_mean          | -3.43e+04    |\n",
            "| time/                   |              |\n",
            "|    fps                  | 88           |\n",
            "|    iterations           | 2            |\n",
            "|    time_elapsed         | 46           |\n",
            "|    total_timesteps      | 4096         |\n",
            "| train/                  |              |\n",
            "|    approx_kl            | 0.0080166655 |\n",
            "|    clip_fraction        | 0.0989       |\n",
            "|    clip_range           | 0.2          |\n",
            "|    entropy_loss         | -5.25        |\n",
            "|    explained_variance   | -0.000336    |\n",
            "|    learning_rate        | 0.001        |\n",
            "|    loss                 | 2.13e+05     |\n",
            "|    n_updates            | 10           |\n",
            "|    policy_gradient_loss | -0.0045      |\n",
            "|    value_loss           | 3.54e+05     |\n",
            "------------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.17e+03    |\n",
            "|    ep_rew_mean          | -3.96e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 88          |\n",
            "|    iterations           | 3           |\n",
            "|    time_elapsed         | 69          |\n",
            "|    total_timesteps      | 6144        |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.007533403 |\n",
            "|    clip_fraction        | 0.0753      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -5.25       |\n",
            "|    explained_variance   | 7.7e-05     |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 1.47e+05    |\n",
            "|    n_updates            | 20          |\n",
            "|    policy_gradient_loss | -0.00421    |\n",
            "|    value_loss           | 3.26e+05    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.17e+03    |\n",
            "|    ep_rew_mean          | -3.98e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 87          |\n",
            "|    iterations           | 4           |\n",
            "|    time_elapsed         | 93          |\n",
            "|    total_timesteps      | 8192        |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.010033689 |\n",
            "|    clip_fraction        | 0.0719      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -5.24       |\n",
            "|    explained_variance   | 2.66e-05    |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 1.95e+05    |\n",
            "|    n_updates            | 30          |\n",
            "|    policy_gradient_loss | -0.003      |\n",
            "|    value_loss           | 4.2e+05     |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.17e+03    |\n",
            "|    ep_rew_mean          | -3.82e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 87          |\n",
            "|    iterations           | 5           |\n",
            "|    time_elapsed         | 117         |\n",
            "|    total_timesteps      | 10240       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.012818455 |\n",
            "|    clip_fraction        | 0.0915      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -5.22       |\n",
            "|    explained_variance   | 7.69e-06    |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 2.01e+05    |\n",
            "|    n_updates            | 40          |\n",
            "|    policy_gradient_loss | -0.00421    |\n",
            "|    value_loss           | 3.99e+05    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.17e+03    |\n",
            "|    ep_rew_mean          | -3.76e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 87          |\n",
            "|    iterations           | 6           |\n",
            "|    time_elapsed         | 140         |\n",
            "|    total_timesteps      | 12288       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.007666164 |\n",
            "|    clip_fraction        | 0.103       |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -5.21       |\n",
            "|    explained_variance   | 1.26e-05    |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 1.02e+05    |\n",
            "|    n_updates            | 50          |\n",
            "|    policy_gradient_loss | -0.00674    |\n",
            "|    value_loss           | 2.09e+05    |\n",
            "-----------------------------------------\n",
            "----------------------------------------\n",
            "| rollout/                |            |\n",
            "|    ep_len_mean          | 1.16e+03   |\n",
            "|    ep_rew_mean          | -3.63e+04  |\n",
            "| time/                   |            |\n",
            "|    fps                  | 87         |\n",
            "|    iterations           | 7          |\n",
            "|    time_elapsed         | 164        |\n",
            "|    total_timesteps      | 14336      |\n",
            "| train/                  |            |\n",
            "|    approx_kl            | 0.00875028 |\n",
            "|    clip_fraction        | 0.0541     |\n",
            "|    clip_range           | 0.2        |\n",
            "|    entropy_loss         | -5.21      |\n",
            "|    explained_variance   | 5.54e-06   |\n",
            "|    learning_rate        | 0.001      |\n",
            "|    loss                 | 1.28e+05   |\n",
            "|    n_updates            | 60         |\n",
            "|    policy_gradient_loss | -0.00115   |\n",
            "|    value_loss           | 3.05e+05   |\n",
            "----------------------------------------\n",
            "------------------------------------------\n",
            "| rollout/                |              |\n",
            "|    ep_len_mean          | 1.14e+03     |\n",
            "|    ep_rew_mean          | -3.45e+04    |\n",
            "| time/                   |              |\n",
            "|    fps                  | 86           |\n",
            "|    iterations           | 8            |\n",
            "|    time_elapsed         | 188          |\n",
            "|    total_timesteps      | 16384        |\n",
            "| train/                  |              |\n",
            "|    approx_kl            | 0.0073153004 |\n",
            "|    clip_fraction        | 0.0674       |\n",
            "|    clip_range           | 0.2          |\n",
            "|    entropy_loss         | -5.19        |\n",
            "|    explained_variance   | 6.79e-06     |\n",
            "|    learning_rate        | 0.001        |\n",
            "|    loss                 | 8.7e+04      |\n",
            "|    n_updates            | 70           |\n",
            "|    policy_gradient_loss | -0.00257     |\n",
            "|    value_loss           | 2.25e+05     |\n",
            "------------------------------------------\n",
            "------------------------------------------\n",
            "| rollout/                |              |\n",
            "|    ep_len_mean          | 1.13e+03     |\n",
            "|    ep_rew_mean          | -3.39e+04    |\n",
            "| time/                   |              |\n",
            "|    fps                  | 87           |\n",
            "|    iterations           | 9            |\n",
            "|    time_elapsed         | 211          |\n",
            "|    total_timesteps      | 18432        |\n",
            "| train/                  |              |\n",
            "|    approx_kl            | 0.0089076115 |\n",
            "|    clip_fraction        | 0.093        |\n",
            "|    clip_range           | 0.2          |\n",
            "|    entropy_loss         | -5.16        |\n",
            "|    explained_variance   | 3.1e-06      |\n",
            "|    learning_rate        | 0.001        |\n",
            "|    loss                 | 8.5e+04      |\n",
            "|    n_updates            | 80           |\n",
            "|    policy_gradient_loss | -0.00341     |\n",
            "|    value_loss           | 1.84e+05     |\n",
            "------------------------------------------\n",
            "------------------------------------------\n",
            "| rollout/                |              |\n",
            "|    ep_len_mean          | 1.11e+03     |\n",
            "|    ep_rew_mean          | -3.21e+04    |\n",
            "| time/                   |              |\n",
            "|    fps                  | 86           |\n",
            "|    iterations           | 10           |\n",
            "|    time_elapsed         | 235          |\n",
            "|    total_timesteps      | 20480        |\n",
            "| train/                  |              |\n",
            "|    approx_kl            | 0.0065261177 |\n",
            "|    clip_fraction        | 0.0747       |\n",
            "|    clip_range           | 0.2          |\n",
            "|    entropy_loss         | -5.14        |\n",
            "|    explained_variance   | 2.8e-06      |\n",
            "|    learning_rate        | 0.001        |\n",
            "|    loss                 | 1.12e+05     |\n",
            "|    n_updates            | 90           |\n",
            "|    policy_gradient_loss | -0.00376     |\n",
            "|    value_loss           | 2.26e+05     |\n",
            "------------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.12e+03    |\n",
            "|    ep_rew_mean          | -3.17e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 86          |\n",
            "|    iterations           | 11          |\n",
            "|    time_elapsed         | 260         |\n",
            "|    total_timesteps      | 22528       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.006153127 |\n",
            "|    clip_fraction        | 0.0759      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -5.12       |\n",
            "|    explained_variance   | 3.7e-06     |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 4.09e+04    |\n",
            "|    n_updates            | 100         |\n",
            "|    policy_gradient_loss | -0.00192    |\n",
            "|    value_loss           | 9.74e+04    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.15e+03    |\n",
            "|    ep_rew_mean          | -3.13e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 86          |\n",
            "|    iterations           | 12          |\n",
            "|    time_elapsed         | 284         |\n",
            "|    total_timesteps      | 24576       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.006908434 |\n",
            "|    clip_fraction        | 0.0933      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -5.09       |\n",
            "|    explained_variance   | 2.21e-06    |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 5.61e+04    |\n",
            "|    n_updates            | 110         |\n",
            "|    policy_gradient_loss | -0.00542    |\n",
            "|    value_loss           | 1.26e+05    |\n",
            "-----------------------------------------\n",
            "------------------------------------------\n",
            "| rollout/                |              |\n",
            "|    ep_len_mean          | 1.16e+03     |\n",
            "|    ep_rew_mean          | -3.12e+04    |\n",
            "| time/                   |              |\n",
            "|    fps                  | 85           |\n",
            "|    iterations           | 13           |\n",
            "|    time_elapsed         | 311          |\n",
            "|    total_timesteps      | 26624        |\n",
            "| train/                  |              |\n",
            "|    approx_kl            | 0.0091868425 |\n",
            "|    clip_fraction        | 0.0862       |\n",
            "|    clip_range           | 0.2          |\n",
            "|    entropy_loss         | -5.1         |\n",
            "|    explained_variance   | 2.32e-06     |\n",
            "|    learning_rate        | 0.001        |\n",
            "|    loss                 | 3.55e+04     |\n",
            "|    n_updates            | 120          |\n",
            "|    policy_gradient_loss | -0.004       |\n",
            "|    value_loss           | 9.28e+04     |\n",
            "------------------------------------------\n",
            "----------------------------------------\n",
            "| rollout/                |            |\n",
            "|    ep_len_mean          | 1.18e+03   |\n",
            "|    ep_rew_mean          | -3.11e+04  |\n",
            "| time/                   |            |\n",
            "|    fps                  | 85         |\n",
            "|    iterations           | 14         |\n",
            "|    time_elapsed         | 336        |\n",
            "|    total_timesteps      | 28672      |\n",
            "| train/                  |            |\n",
            "|    approx_kl            | 0.00992664 |\n",
            "|    clip_fraction        | 0.118      |\n",
            "|    clip_range           | 0.2        |\n",
            "|    entropy_loss         | -5.11      |\n",
            "|    explained_variance   | 1.79e-06   |\n",
            "|    learning_rate        | 0.001      |\n",
            "|    loss                 | 5.5e+04    |\n",
            "|    n_updates            | 130        |\n",
            "|    policy_gradient_loss | -0.00564   |\n",
            "|    value_loss           | 1.13e+05   |\n",
            "----------------------------------------\n",
            "----------------------------------------\n",
            "| rollout/                |            |\n",
            "|    ep_len_mean          | 1.19e+03   |\n",
            "|    ep_rew_mean          | -3.08e+04  |\n",
            "| time/                   |            |\n",
            "|    fps                  | 84         |\n",
            "|    iterations           | 15         |\n",
            "|    time_elapsed         | 362        |\n",
            "|    total_timesteps      | 30720      |\n",
            "| train/                  |            |\n",
            "|    approx_kl            | 0.00821295 |\n",
            "|    clip_fraction        | 0.0692     |\n",
            "|    clip_range           | 0.2        |\n",
            "|    entropy_loss         | -5.09      |\n",
            "|    explained_variance   | 1.49e-06   |\n",
            "|    learning_rate        | 0.001      |\n",
            "|    loss                 | 6.25e+04   |\n",
            "|    n_updates            | 140        |\n",
            "|    policy_gradient_loss | -0.00351   |\n",
            "|    value_loss           | 1.42e+05   |\n",
            "----------------------------------------\n",
            "------------------------------------------\n",
            "| rollout/                |              |\n",
            "|    ep_len_mean          | 1.19e+03     |\n",
            "|    ep_rew_mean          | -2.99e+04    |\n",
            "| time/                   |              |\n",
            "|    fps                  | 84           |\n",
            "|    iterations           | 16           |\n",
            "|    time_elapsed         | 386          |\n",
            "|    total_timesteps      | 32768        |\n",
            "| train/                  |              |\n",
            "|    approx_kl            | 0.0064286557 |\n",
            "|    clip_fraction        | 0.0776       |\n",
            "|    clip_range           | 0.2          |\n",
            "|    entropy_loss         | -5.1         |\n",
            "|    explained_variance   | 1.55e-06     |\n",
            "|    learning_rate        | 0.001        |\n",
            "|    loss                 | 3.46e+04     |\n",
            "|    n_updates            | 150          |\n",
            "|    policy_gradient_loss | -0.0024      |\n",
            "|    value_loss           | 6.83e+04     |\n",
            "------------------------------------------\n",
            "------------------------------------------\n",
            "| rollout/                |              |\n",
            "|    ep_len_mean          | 1.2e+03      |\n",
            "|    ep_rew_mean          | -2.96e+04    |\n",
            "| time/                   |              |\n",
            "|    fps                  | 85           |\n",
            "|    iterations           | 17           |\n",
            "|    time_elapsed         | 409          |\n",
            "|    total_timesteps      | 34816        |\n",
            "| train/                  |              |\n",
            "|    approx_kl            | 0.0052765226 |\n",
            "|    clip_fraction        | 0.068        |\n",
            "|    clip_range           | 0.2          |\n",
            "|    entropy_loss         | -5.09        |\n",
            "|    explained_variance   | 1.31e-06     |\n",
            "|    learning_rate        | 0.001        |\n",
            "|    loss                 | 2.14e+04     |\n",
            "|    n_updates            | 160          |\n",
            "|    policy_gradient_loss | -0.00285     |\n",
            "|    value_loss           | 4.88e+04     |\n",
            "------------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.21e+03    |\n",
            "|    ep_rew_mean          | -2.93e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 84          |\n",
            "|    iterations           | 18          |\n",
            "|    time_elapsed         | 434         |\n",
            "|    total_timesteps      | 36864       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.009790231 |\n",
            "|    clip_fraction        | 0.0967      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -5.09       |\n",
            "|    explained_variance   | 1.01e-06    |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 3.54e+04    |\n",
            "|    n_updates            | 170         |\n",
            "|    policy_gradient_loss | -0.00654    |\n",
            "|    value_loss           | 7.63e+04    |\n",
            "-----------------------------------------\n",
            "------------------------------------------\n",
            "| rollout/                |              |\n",
            "|    ep_len_mean          | 1.21e+03     |\n",
            "|    ep_rew_mean          | -2.87e+04    |\n",
            "| time/                   |              |\n",
            "|    fps                  | 84           |\n",
            "|    iterations           | 19           |\n",
            "|    time_elapsed         | 460          |\n",
            "|    total_timesteps      | 38912        |\n",
            "| train/                  |              |\n",
            "|    approx_kl            | 0.0056219227 |\n",
            "|    clip_fraction        | 0.0486       |\n",
            "|    clip_range           | 0.2          |\n",
            "|    entropy_loss         | -5.09        |\n",
            "|    explained_variance   | 8.94e-07     |\n",
            "|    learning_rate        | 0.001        |\n",
            "|    loss                 | 2.75e+04     |\n",
            "|    n_updates            | 180          |\n",
            "|    policy_gradient_loss | -0.00206     |\n",
            "|    value_loss           | 5.69e+04     |\n",
            "------------------------------------------\n",
            "------------------------------------------\n",
            "| rollout/                |              |\n",
            "|    ep_len_mean          | 1.24e+03     |\n",
            "|    ep_rew_mean          | -2.83e+04    |\n",
            "| time/                   |              |\n",
            "|    fps                  | 84           |\n",
            "|    iterations           | 20           |\n",
            "|    time_elapsed         | 486          |\n",
            "|    total_timesteps      | 40960        |\n",
            "| train/                  |              |\n",
            "|    approx_kl            | 0.0068632695 |\n",
            "|    clip_fraction        | 0.0707       |\n",
            "|    clip_range           | 0.2          |\n",
            "|    entropy_loss         | -5.07        |\n",
            "|    explained_variance   | 1.01e-06     |\n",
            "|    learning_rate        | 0.001        |\n",
            "|    loss                 | 1.33e+04     |\n",
            "|    n_updates            | 190          |\n",
            "|    policy_gradient_loss | -0.00153     |\n",
            "|    value_loss           | 3.4e+04      |\n",
            "------------------------------------------\n",
            "------------------------------------------\n",
            "| rollout/                |              |\n",
            "|    ep_len_mean          | 1.25e+03     |\n",
            "|    ep_rew_mean          | -2.8e+04     |\n",
            "| time/                   |              |\n",
            "|    fps                  | 83           |\n",
            "|    iterations           | 21           |\n",
            "|    time_elapsed         | 512          |\n",
            "|    total_timesteps      | 43008        |\n",
            "| train/                  |              |\n",
            "|    approx_kl            | 0.0066338973 |\n",
            "|    clip_fraction        | 0.0757       |\n",
            "|    clip_range           | 0.2          |\n",
            "|    entropy_loss         | -5.05        |\n",
            "|    explained_variance   | 1.19e-06     |\n",
            "|    learning_rate        | 0.001        |\n",
            "|    loss                 | 1.25e+04     |\n",
            "|    n_updates            | 200          |\n",
            "|    policy_gradient_loss | -0.00374     |\n",
            "|    value_loss           | 2.58e+04     |\n",
            "------------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.25e+03    |\n",
            "|    ep_rew_mean          | -2.77e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 22          |\n",
            "|    time_elapsed         | 536         |\n",
            "|    total_timesteps      | 45056       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.009644468 |\n",
            "|    clip_fraction        | 0.133       |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -5.01       |\n",
            "|    explained_variance   | 7.15e-07    |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 5.28e+03    |\n",
            "|    n_updates            | 210         |\n",
            "|    policy_gradient_loss | -0.006      |\n",
            "|    value_loss           | 1.68e+04    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.26e+03    |\n",
            "|    ep_rew_mean          | -2.7e+04    |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 23          |\n",
            "|    time_elapsed         | 563         |\n",
            "|    total_timesteps      | 47104       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.011505453 |\n",
            "|    clip_fraction        | 0.132       |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -5.02       |\n",
            "|    explained_variance   | 6.56e-07    |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 1.03e+04    |\n",
            "|    n_updates            | 220         |\n",
            "|    policy_gradient_loss | -0.00895    |\n",
            "|    value_loss           | 2.2e+04     |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.28e+03    |\n",
            "|    ep_rew_mean          | -2.67e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 24          |\n",
            "|    time_elapsed         | 587         |\n",
            "|    total_timesteps      | 49152       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.008028438 |\n",
            "|    clip_fraction        | 0.102       |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -5.01       |\n",
            "|    explained_variance   | 8.34e-07    |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 3.41e+03    |\n",
            "|    n_updates            | 230         |\n",
            "|    policy_gradient_loss | -0.00221    |\n",
            "|    value_loss           | 1.17e+04    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.28e+03    |\n",
            "|    ep_rew_mean          | -2.63e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 25          |\n",
            "|    time_elapsed         | 611         |\n",
            "|    total_timesteps      | 51200       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.007652807 |\n",
            "|    clip_fraction        | 0.0664      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.96       |\n",
            "|    explained_variance   | 2.98e-07    |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 3.69e+03    |\n",
            "|    n_updates            | 240         |\n",
            "|    policy_gradient_loss | -0.000185   |\n",
            "|    value_loss           | 8.9e+03     |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.29e+03    |\n",
            "|    ep_rew_mean          | -2.56e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 26          |\n",
            "|    time_elapsed         | 636         |\n",
            "|    total_timesteps      | 53248       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.011867992 |\n",
            "|    clip_fraction        | 0.106       |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.94       |\n",
            "|    explained_variance   | 5.96e-08    |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 6.38e+03    |\n",
            "|    n_updates            | 250         |\n",
            "|    policy_gradient_loss | -0.00248    |\n",
            "|    value_loss           | 8.41e+03    |\n",
            "-----------------------------------------\n",
            "------------------------------------------\n",
            "| rollout/                |              |\n",
            "|    ep_len_mean          | 1.3e+03      |\n",
            "|    ep_rew_mean          | -2.52e+04    |\n",
            "| time/                   |              |\n",
            "|    fps                  | 83           |\n",
            "|    iterations           | 27           |\n",
            "|    time_elapsed         | 661          |\n",
            "|    total_timesteps      | 55296        |\n",
            "| train/                  |              |\n",
            "|    approx_kl            | 0.0075610094 |\n",
            "|    clip_fraction        | 0.353        |\n",
            "|    clip_range           | 0.2          |\n",
            "|    entropy_loss         | -4.86        |\n",
            "|    explained_variance   | 4.17e-07     |\n",
            "|    learning_rate        | 0.001        |\n",
            "|    loss                 | 3.22e+03     |\n",
            "|    n_updates            | 260          |\n",
            "|    policy_gradient_loss | 0.0259       |\n",
            "|    value_loss           | 4.57e+03     |\n",
            "------------------------------------------\n",
            "------------------------------------------\n",
            "| rollout/                |              |\n",
            "|    ep_len_mean          | 1.3e+03      |\n",
            "|    ep_rew_mean          | -2.46e+04    |\n",
            "| time/                   |              |\n",
            "|    fps                  | 83           |\n",
            "|    iterations           | 28           |\n",
            "|    time_elapsed         | 685          |\n",
            "|    total_timesteps      | 57344        |\n",
            "| train/                  |              |\n",
            "|    approx_kl            | 0.0056704506 |\n",
            "|    clip_fraction        | 0.0424       |\n",
            "|    clip_range           | 0.2          |\n",
            "|    entropy_loss         | -4.94        |\n",
            "|    explained_variance   | 0.192        |\n",
            "|    learning_rate        | 0.001        |\n",
            "|    loss                 | 695          |\n",
            "|    n_updates            | 270          |\n",
            "|    policy_gradient_loss | -0.00454     |\n",
            "|    value_loss           | 2.97e+03     |\n",
            "------------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.31e+03    |\n",
            "|    ep_rew_mean          | -2.43e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 29          |\n",
            "|    time_elapsed         | 711         |\n",
            "|    total_timesteps      | 59392       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.006635526 |\n",
            "|    clip_fraction        | 0.0538      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.91       |\n",
            "|    explained_variance   | 0.604       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 5.58e+03    |\n",
            "|    n_updates            | 280         |\n",
            "|    policy_gradient_loss | -0.00402    |\n",
            "|    value_loss           | 5.84e+03    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.31e+03    |\n",
            "|    ep_rew_mean          | -2.4e+04    |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 30          |\n",
            "|    time_elapsed         | 736         |\n",
            "|    total_timesteps      | 61440       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.006966117 |\n",
            "|    clip_fraction        | 0.0962      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.83       |\n",
            "|    explained_variance   | 0.664       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 268         |\n",
            "|    n_updates            | 290         |\n",
            "|    policy_gradient_loss | -0.00388    |\n",
            "|    value_loss           | 1.57e+03    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.32e+03    |\n",
            "|    ep_rew_mean          | -2.34e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 31          |\n",
            "|    time_elapsed         | 761         |\n",
            "|    total_timesteps      | 63488       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.005343941 |\n",
            "|    clip_fraction        | 0.105       |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.8        |\n",
            "|    explained_variance   | 0.712       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 60.7        |\n",
            "|    n_updates            | 300         |\n",
            "|    policy_gradient_loss | 0.000694    |\n",
            "|    value_loss           | 1.44e+03    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.31e+03    |\n",
            "|    ep_rew_mean          | -2.31e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 32          |\n",
            "|    time_elapsed         | 786         |\n",
            "|    total_timesteps      | 65536       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.010675106 |\n",
            "|    clip_fraction        | 0.14        |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.86       |\n",
            "|    explained_variance   | 0.62        |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 346         |\n",
            "|    n_updates            | 310         |\n",
            "|    policy_gradient_loss | -0.00467    |\n",
            "|    value_loss           | 3.1e+03     |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.33e+03    |\n",
            "|    ep_rew_mean          | -2.29e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 33          |\n",
            "|    time_elapsed         | 811         |\n",
            "|    total_timesteps      | 67584       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.009389532 |\n",
            "|    clip_fraction        | 0.113       |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.8        |\n",
            "|    explained_variance   | 0.748       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 1.13e+03    |\n",
            "|    n_updates            | 320         |\n",
            "|    policy_gradient_loss | -0.00691    |\n",
            "|    value_loss           | 3.42e+03    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.33e+03    |\n",
            "|    ep_rew_mean          | -2.25e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 34          |\n",
            "|    time_elapsed         | 834         |\n",
            "|    total_timesteps      | 69632       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.006670535 |\n",
            "|    clip_fraction        | 0.0528      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.81       |\n",
            "|    explained_variance   | 0.606       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 976         |\n",
            "|    n_updates            | 330         |\n",
            "|    policy_gradient_loss | -0.00274    |\n",
            "|    value_loss           | 3.26e+03    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.33e+03    |\n",
            "|    ep_rew_mean          | -2.22e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 35          |\n",
            "|    time_elapsed         | 857         |\n",
            "|    total_timesteps      | 71680       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.008612213 |\n",
            "|    clip_fraction        | 0.0987      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.78       |\n",
            "|    explained_variance   | 0.664       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 1.22e+03    |\n",
            "|    n_updates            | 340         |\n",
            "|    policy_gradient_loss | -0.00427    |\n",
            "|    value_loss           | 5.8e+03     |\n",
            "-----------------------------------------\n",
            "----------------------------------------\n",
            "| rollout/                |            |\n",
            "|    ep_len_mean          | 1.33e+03   |\n",
            "|    ep_rew_mean          | -2.19e+04  |\n",
            "| time/                   |            |\n",
            "|    fps                  | 83         |\n",
            "|    iterations           | 36         |\n",
            "|    time_elapsed         | 881        |\n",
            "|    total_timesteps      | 73728      |\n",
            "| train/                  |            |\n",
            "|    approx_kl            | 0.00791744 |\n",
            "|    clip_fraction        | 0.12       |\n",
            "|    clip_range           | 0.2        |\n",
            "|    entropy_loss         | -4.77      |\n",
            "|    explained_variance   | 0.79       |\n",
            "|    learning_rate        | 0.001      |\n",
            "|    loss                 | 160        |\n",
            "|    n_updates            | 350        |\n",
            "|    policy_gradient_loss | -0.0045    |\n",
            "|    value_loss           | 2.02e+03   |\n",
            "----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.32e+03    |\n",
            "|    ep_rew_mean          | -2.13e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 37          |\n",
            "|    time_elapsed         | 906         |\n",
            "|    total_timesteps      | 75776       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.012531351 |\n",
            "|    clip_fraction        | 0.107       |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.8        |\n",
            "|    explained_variance   | 0.263       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 1.77e+03    |\n",
            "|    n_updates            | 360         |\n",
            "|    policy_gradient_loss | -0.00511    |\n",
            "|    value_loss           | 5.47e+03    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.33e+03    |\n",
            "|    ep_rew_mean          | -2.11e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 38          |\n",
            "|    time_elapsed         | 933         |\n",
            "|    total_timesteps      | 77824       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.008859118 |\n",
            "|    clip_fraction        | 0.161       |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.76       |\n",
            "|    explained_variance   | 0.688       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 3.18e+03    |\n",
            "|    n_updates            | 370         |\n",
            "|    policy_gradient_loss | -0.00285    |\n",
            "|    value_loss           | 3.47e+03    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.34e+03    |\n",
            "|    ep_rew_mean          | -2.08e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 39          |\n",
            "|    time_elapsed         | 960         |\n",
            "|    total_timesteps      | 79872       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.011655061 |\n",
            "|    clip_fraction        | 0.131       |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.72       |\n",
            "|    explained_variance   | 0.838       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 3.67e+03    |\n",
            "|    n_updates            | 380         |\n",
            "|    policy_gradient_loss | 0.00183     |\n",
            "|    value_loss           | 1.65e+03    |\n",
            "-----------------------------------------\n",
            "----------------------------------------\n",
            "| rollout/                |            |\n",
            "|    ep_len_mean          | 1.35e+03   |\n",
            "|    ep_rew_mean          | -2.06e+04  |\n",
            "| time/                   |            |\n",
            "|    fps                  | 83         |\n",
            "|    iterations           | 40         |\n",
            "|    time_elapsed         | 984        |\n",
            "|    total_timesteps      | 81920      |\n",
            "| train/                  |            |\n",
            "|    approx_kl            | 0.01071727 |\n",
            "|    clip_fraction        | 0.206      |\n",
            "|    clip_range           | 0.2        |\n",
            "|    entropy_loss         | -4.71      |\n",
            "|    explained_variance   | 0.602      |\n",
            "|    learning_rate        | 0.001      |\n",
            "|    loss                 | 190        |\n",
            "|    n_updates            | 390        |\n",
            "|    policy_gradient_loss | 0.00763    |\n",
            "|    value_loss           | 1.45e+03   |\n",
            "----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.36e+03    |\n",
            "|    ep_rew_mean          | -2.04e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 41          |\n",
            "|    time_elapsed         | 1008        |\n",
            "|    total_timesteps      | 83968       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.010233748 |\n",
            "|    clip_fraction        | 0.134       |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.74       |\n",
            "|    explained_variance   | 0.647       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 1.94e+03    |\n",
            "|    n_updates            | 400         |\n",
            "|    policy_gradient_loss | 0.00158     |\n",
            "|    value_loss           | 1.27e+03    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.36e+03    |\n",
            "|    ep_rew_mean          | -2.02e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 42          |\n",
            "|    time_elapsed         | 1031        |\n",
            "|    total_timesteps      | 86016       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.011342964 |\n",
            "|    clip_fraction        | 0.18        |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.63       |\n",
            "|    explained_variance   | 0.752       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 22.6        |\n",
            "|    n_updates            | 410         |\n",
            "|    policy_gradient_loss | -0.000163   |\n",
            "|    value_loss           | 1.18e+03    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.37e+03    |\n",
            "|    ep_rew_mean          | -2e+04      |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 43          |\n",
            "|    time_elapsed         | 1055        |\n",
            "|    total_timesteps      | 88064       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.010002422 |\n",
            "|    clip_fraction        | 0.0957      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.74       |\n",
            "|    explained_variance   | 0.704       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 293         |\n",
            "|    n_updates            | 420         |\n",
            "|    policy_gradient_loss | -0.00397    |\n",
            "|    value_loss           | 1.94e+03    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.37e+03    |\n",
            "|    ep_rew_mean          | -1.98e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 44          |\n",
            "|    time_elapsed         | 1080        |\n",
            "|    total_timesteps      | 90112       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.009995557 |\n",
            "|    clip_fraction        | 0.0781      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.74       |\n",
            "|    explained_variance   | 0.717       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 2.3e+03     |\n",
            "|    n_updates            | 430         |\n",
            "|    policy_gradient_loss | -0.00329    |\n",
            "|    value_loss           | 3.07e+03    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.37e+03    |\n",
            "|    ep_rew_mean          | -1.95e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 45          |\n",
            "|    time_elapsed         | 1105        |\n",
            "|    total_timesteps      | 92160       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.009772761 |\n",
            "|    clip_fraction        | 0.0829      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.71       |\n",
            "|    explained_variance   | 0.279       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 374         |\n",
            "|    n_updates            | 440         |\n",
            "|    policy_gradient_loss | -0.00165    |\n",
            "|    value_loss           | 2.34e+03    |\n",
            "-----------------------------------------\n",
            "------------------------------------------\n",
            "| rollout/                |              |\n",
            "|    ep_len_mean          | 1.37e+03     |\n",
            "|    ep_rew_mean          | -1.94e+04    |\n",
            "| time/                   |              |\n",
            "|    fps                  | 83           |\n",
            "|    iterations           | 46           |\n",
            "|    time_elapsed         | 1131         |\n",
            "|    total_timesteps      | 94208        |\n",
            "| train/                  |              |\n",
            "|    approx_kl            | 0.0062507237 |\n",
            "|    clip_fraction        | 0.0695       |\n",
            "|    clip_range           | 0.2          |\n",
            "|    entropy_loss         | -4.76        |\n",
            "|    explained_variance   | 0.461        |\n",
            "|    learning_rate        | 0.001        |\n",
            "|    loss                 | 343          |\n",
            "|    n_updates            | 450          |\n",
            "|    policy_gradient_loss | -0.00568     |\n",
            "|    value_loss           | 3.53e+03     |\n",
            "------------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.38e+03    |\n",
            "|    ep_rew_mean          | -1.93e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 83          |\n",
            "|    iterations           | 47          |\n",
            "|    time_elapsed         | 1158        |\n",
            "|    total_timesteps      | 96256       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.008773536 |\n",
            "|    clip_fraction        | 0.0684      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.73       |\n",
            "|    explained_variance   | 0.667       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 309         |\n",
            "|    n_updates            | 460         |\n",
            "|    policy_gradient_loss | -0.00211    |\n",
            "|    value_loss           | 1.96e+03    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.38e+03    |\n",
            "|    ep_rew_mean          | -1.89e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 82          |\n",
            "|    iterations           | 48          |\n",
            "|    time_elapsed         | 1189        |\n",
            "|    total_timesteps      | 98304       |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.014286237 |\n",
            "|    clip_fraction        | 0.114       |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.71       |\n",
            "|    explained_variance   | 0.747       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 135         |\n",
            "|    n_updates            | 470         |\n",
            "|    policy_gradient_loss | -0.00267    |\n",
            "|    value_loss           | 1.8e+03     |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.39e+03    |\n",
            "|    ep_rew_mean          | -1.88e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 82          |\n",
            "|    iterations           | 49          |\n",
            "|    time_elapsed         | 1216        |\n",
            "|    total_timesteps      | 100352      |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.009100048 |\n",
            "|    clip_fraction        | 0.0908      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.73       |\n",
            "|    explained_variance   | 0.523       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 1.07e+03    |\n",
            "|    n_updates            | 480         |\n",
            "|    policy_gradient_loss | -0.00154    |\n",
            "|    value_loss           | 3.45e+03    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.39e+03    |\n",
            "|    ep_rew_mean          | -1.87e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 82          |\n",
            "|    iterations           | 50          |\n",
            "|    time_elapsed         | 1242        |\n",
            "|    total_timesteps      | 102400      |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.006386128 |\n",
            "|    clip_fraction        | 0.0952      |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.71       |\n",
            "|    explained_variance   | 0.758       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 87.4        |\n",
            "|    n_updates            | 490         |\n",
            "|    policy_gradient_loss | -0.00293    |\n",
            "|    value_loss           | 1.39e+03    |\n",
            "-----------------------------------------\n",
            "-----------------------------------------\n",
            "| rollout/                |             |\n",
            "|    ep_len_mean          | 1.39e+03    |\n",
            "|    ep_rew_mean          | -1.86e+04   |\n",
            "| time/                   |             |\n",
            "|    fps                  | 82          |\n",
            "|    iterations           | 51          |\n",
            "|    time_elapsed         | 1266        |\n",
            "|    total_timesteps      | 104448      |\n",
            "| train/                  |             |\n",
            "|    approx_kl            | 0.009490039 |\n",
            "|    clip_fraction        | 0.11        |\n",
            "|    clip_range           | 0.2         |\n",
            "|    entropy_loss         | -4.66       |\n",
            "|    explained_variance   | 0.787       |\n",
            "|    learning_rate        | 0.001       |\n",
            "|    loss                 | 683         |\n",
            "|    n_updates            | 500         |\n",
            "|    policy_gradient_loss | -0.00605    |\n",
            "|    value_loss           | 1.56e+03    |\n",
            "-----------------------------------------\n",
            "------------------------------------------\n",
            "| rollout/                |              |\n",
            "|    ep_len_mean          | 1.39e+03     |\n",
            "|    ep_rew_mean          | -1.83e+04    |\n",
            "| time/                   |              |\n",
            "|    fps                  | 82           |\n",
            "|    iterations           | 52           |\n",
            "|    time_elapsed         | 1293         |\n",
            "|    total_timesteps      | 106496       |\n",
            "| train/                  |              |\n",
            "|    approx_kl            | 0.0115320925 |\n",
            "|    clip_fraction        | 0.0973       |\n",
            "|    clip_range           | 0.2          |\n",
            "|    entropy_loss         | -4.66        |\n",
            "|    explained_variance   | 0.859        |\n",
            "|    learning_rate        | 0.001        |\n",
            "|    loss                 | 281          |\n",
            "|    n_updates            | 510          |\n",
            "|    policy_gradient_loss | -0.00405     |\n",
            "|    value_loss           | 2.05e+03     |\n",
            "------------------------------------------\n",
            "------------------------------------------\n",
            "| rollout/                |              |\n",
            "|    ep_len_mean          | 1.39e+03     |\n",
            "|    ep_rew_mean          | -1.81e+04    |\n",
            "| time/                   |              |\n",
            "|    fps                  | 82           |\n",
            "|    iterations           | 53           |\n",
            "|    time_elapsed         | 1319         |\n",
            "|    total_timesteps      | 108544       |\n",
            "| train/                  |              |\n",
            "|    approx_kl            | 0.0074732862 |\n",
            "|    clip_fraction        | 0.127        |\n",
            "|    clip_range           | 0.2          |\n",
            "|    entropy_loss         | -4.73        |\n",
            "|    explained_variance   | 0.776        |\n",
            "|    learning_rate        | 0.001        |\n",
            "|    loss                 | 642          |\n",
            "|    n_updates            | 520          |\n",
            "|    policy_gradient_loss | -0.00489     |\n",
            "|    value_loss           | 2.99e+03     |\n",
            "------------------------------------------\n"
          ]
        },
        {
          "ename": "KeyboardInterrupt",
          "evalue": "",
          "output_type": "error",
          "traceback": [
            "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
            "\u001b[1;32mg:\\git_projects\\Battery-RL\\LIB_DQN.ipynb Cell 9\u001b[0m line \u001b[0;36m3\n\u001b[0;32m     <a href='vscode-notebook-cell:/g%3A/git_projects/Battery-RL/LIB_DQN.ipynb#X10sZmlsZQ%3D%3D?line=25'>26</a>\u001b[0m callback \u001b[39m=\u001b[39m SaveOnBestTrainingRewardCallback(check_freq\u001b[39m=\u001b[39m\u001b[39m2e4\u001b[39m, log_dir\u001b[39m=\u001b[39mlog_dir)\n\u001b[0;32m     <a href='vscode-notebook-cell:/g%3A/git_projects/Battery-RL/LIB_DQN.ipynb#X10sZmlsZQ%3D%3D?line=27'>28</a>\u001b[0m \u001b[39m# Train the agent\u001b[39;00m\n\u001b[0;32m     <a href='vscode-notebook-cell:/g%3A/git_projects/Battery-RL/LIB_DQN.ipynb#X10sZmlsZQ%3D%3D?line=28'>29</a>\u001b[0m \u001b[39m# model = model.load('best_model.zip', env)\u001b[39;00m\n\u001b[1;32m---> <a href='vscode-notebook-cell:/g%3A/git_projects/Battery-RL/LIB_DQN.ipynb#X10sZmlsZQ%3D%3D?line=29'>30</a>\u001b[0m model\u001b[39m.\u001b[39;49mlearn(total_timesteps\u001b[39m=\u001b[39;49mtimesteps, progress_bar\u001b[39m=\u001b[39;49m\u001b[39mFalse\u001b[39;49;00m)\n",
            "File \u001b[1;32md:\\Users\\amirh\\anaconda3\\lib\\site-packages\\stable_baselines3\\ppo\\ppo.py:308\u001b[0m, in \u001b[0;36mPPO.learn\u001b[1;34m(self, total_timesteps, callback, log_interval, tb_log_name, reset_num_timesteps, progress_bar)\u001b[0m\n\u001b[0;32m    299\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39mlearn\u001b[39m(\n\u001b[0;32m    300\u001b[0m     \u001b[39mself\u001b[39m: SelfPPO,\n\u001b[0;32m    301\u001b[0m     total_timesteps: \u001b[39mint\u001b[39m,\n\u001b[1;32m   (...)\u001b[0m\n\u001b[0;32m    306\u001b[0m     progress_bar: \u001b[39mbool\u001b[39m \u001b[39m=\u001b[39m \u001b[39mFalse\u001b[39;00m,\n\u001b[0;32m    307\u001b[0m ) \u001b[39m-\u001b[39m\u001b[39m>\u001b[39m SelfPPO:\n\u001b[1;32m--> 308\u001b[0m     \u001b[39mreturn\u001b[39;00m \u001b[39msuper\u001b[39;49m()\u001b[39m.\u001b[39;49mlearn(\n\u001b[0;32m    309\u001b[0m         total_timesteps\u001b[39m=\u001b[39;49mtotal_timesteps,\n\u001b[0;32m    310\u001b[0m         callback\u001b[39m=\u001b[39;49mcallback,\n\u001b[0;32m    311\u001b[0m         log_interval\u001b[39m=\u001b[39;49mlog_interval,\n\u001b[0;32m    312\u001b[0m         tb_log_name\u001b[39m=\u001b[39;49mtb_log_name,\n\u001b[0;32m    313\u001b[0m         reset_num_timesteps\u001b[39m=\u001b[39;49mreset_num_timesteps,\n\u001b[0;32m    314\u001b[0m         progress_bar\u001b[39m=\u001b[39;49mprogress_bar,\n\u001b[0;32m    315\u001b[0m     )\n",
            "File \u001b[1;32md:\\Users\\amirh\\anaconda3\\lib\\site-packages\\stable_baselines3\\common\\on_policy_algorithm.py:259\u001b[0m, in \u001b[0;36mOnPolicyAlgorithm.learn\u001b[1;34m(self, total_timesteps, callback, log_interval, tb_log_name, reset_num_timesteps, progress_bar)\u001b[0m\n\u001b[0;32m    256\u001b[0m \u001b[39massert\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39menv \u001b[39mis\u001b[39;00m \u001b[39mnot\u001b[39;00m \u001b[39mNone\u001b[39;00m\n\u001b[0;32m    258\u001b[0m \u001b[39mwhile\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mnum_timesteps \u001b[39m<\u001b[39m total_timesteps:\n\u001b[1;32m--> 259\u001b[0m     continue_training \u001b[39m=\u001b[39m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mcollect_rollouts(\u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49menv, callback, \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mrollout_buffer, n_rollout_steps\u001b[39m=\u001b[39;49m\u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mn_steps)\n\u001b[0;32m    261\u001b[0m     \u001b[39mif\u001b[39;00m continue_training \u001b[39mis\u001b[39;00m \u001b[39mFalse\u001b[39;00m:\n\u001b[0;32m    262\u001b[0m         \u001b[39mbreak\u001b[39;00m\n",
            "File \u001b[1;32md:\\Users\\amirh\\anaconda3\\lib\\site-packages\\stable_baselines3\\common\\on_policy_algorithm.py:169\u001b[0m, in \u001b[0;36mOnPolicyAlgorithm.collect_rollouts\u001b[1;34m(self, env, callback, rollout_buffer, n_rollout_steps)\u001b[0m\n\u001b[0;32m    166\u001b[0m \u001b[39mwith\u001b[39;00m th\u001b[39m.\u001b[39mno_grad():\n\u001b[0;32m    167\u001b[0m     \u001b[39m# Convert to pytorch tensor or to TensorDict\u001b[39;00m\n\u001b[0;32m    168\u001b[0m     obs_tensor \u001b[39m=\u001b[39m obs_as_tensor(\u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_last_obs, \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mdevice)\n\u001b[1;32m--> 169\u001b[0m     actions, values, log_probs \u001b[39m=\u001b[39m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mpolicy(obs_tensor)\n\u001b[0;32m    170\u001b[0m actions \u001b[39m=\u001b[39m actions\u001b[39m.\u001b[39mcpu()\u001b[39m.\u001b[39mnumpy()\n\u001b[0;32m    172\u001b[0m \u001b[39m# Rescale and perform action\u001b[39;00m\n",
            "File \u001b[1;32md:\\Users\\amirh\\anaconda3\\lib\\site-packages\\torch\\nn\\modules\\module.py:1501\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[1;34m(self, *args, **kwargs)\u001b[0m\n\u001b[0;32m   1496\u001b[0m \u001b[39m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[0;32m   1497\u001b[0m \u001b[39m# this function, and just call forward.\u001b[39;00m\n\u001b[0;32m   1498\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mnot\u001b[39;00m (\u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_backward_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_backward_pre_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_forward_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_forward_pre_hooks\n\u001b[0;32m   1499\u001b[0m         \u001b[39mor\u001b[39;00m _global_backward_pre_hooks \u001b[39mor\u001b[39;00m _global_backward_hooks\n\u001b[0;32m   1500\u001b[0m         \u001b[39mor\u001b[39;00m _global_forward_hooks \u001b[39mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[1;32m-> 1501\u001b[0m     \u001b[39mreturn\u001b[39;00m forward_call(\u001b[39m*\u001b[39margs, \u001b[39m*\u001b[39m\u001b[39m*\u001b[39mkwargs)\n\u001b[0;32m   1502\u001b[0m \u001b[39m# Do not call functions when jit is used\u001b[39;00m\n\u001b[0;32m   1503\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[39m=\u001b[39m [], []\n",
            "File \u001b[1;32md:\\Users\\amirh\\anaconda3\\lib\\site-packages\\stable_baselines3\\common\\policies.py:626\u001b[0m, in \u001b[0;36mActorCriticPolicy.forward\u001b[1;34m(self, obs, deterministic)\u001b[0m\n\u001b[0;32m    624\u001b[0m \u001b[39m# Evaluate the values for the given observations\u001b[39;00m\n\u001b[0;32m    625\u001b[0m values \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mvalue_net(latent_vf)\n\u001b[1;32m--> 626\u001b[0m distribution \u001b[39m=\u001b[39m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_get_action_dist_from_latent(latent_pi)\n\u001b[0;32m    627\u001b[0m actions \u001b[39m=\u001b[39m distribution\u001b[39m.\u001b[39mget_actions(deterministic\u001b[39m=\u001b[39mdeterministic)\n\u001b[0;32m    628\u001b[0m log_prob \u001b[39m=\u001b[39m distribution\u001b[39m.\u001b[39mlog_prob(actions)\n",
            "File \u001b[1;32md:\\Users\\amirh\\anaconda3\\lib\\site-packages\\stable_baselines3\\common\\policies.py:662\u001b[0m, in \u001b[0;36mActorCriticPolicy._get_action_dist_from_latent\u001b[1;34m(self, latent_pi)\u001b[0m\n\u001b[0;32m    659\u001b[0m     \u001b[39mreturn\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39maction_dist\u001b[39m.\u001b[39mproba_distribution(action_logits\u001b[39m=\u001b[39mmean_actions)\n\u001b[0;32m    660\u001b[0m \u001b[39melif\u001b[39;00m \u001b[39misinstance\u001b[39m(\u001b[39mself\u001b[39m\u001b[39m.\u001b[39maction_dist, MultiCategoricalDistribution):\n\u001b[0;32m    661\u001b[0m     \u001b[39m# Here mean_actions are the flattened logits\u001b[39;00m\n\u001b[1;32m--> 662\u001b[0m     \u001b[39mreturn\u001b[39;00m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49maction_dist\u001b[39m.\u001b[39;49mproba_distribution(action_logits\u001b[39m=\u001b[39;49mmean_actions)\n\u001b[0;32m    663\u001b[0m \u001b[39melif\u001b[39;00m \u001b[39misinstance\u001b[39m(\u001b[39mself\u001b[39m\u001b[39m.\u001b[39maction_dist, BernoulliDistribution):\n\u001b[0;32m    664\u001b[0m     \u001b[39m# Here mean_actions are the logits (before rounding to get the binary actions)\u001b[39;00m\n\u001b[0;32m    665\u001b[0m     \u001b[39mreturn\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39maction_dist\u001b[39m.\u001b[39mproba_distribution(action_logits\u001b[39m=\u001b[39mmean_actions)\n",
            "File \u001b[1;32md:\\Users\\amirh\\anaconda3\\lib\\site-packages\\stable_baselines3\\common\\distributions.py:342\u001b[0m, in \u001b[0;36mMultiCategoricalDistribution.proba_distribution\u001b[1;34m(self, action_logits)\u001b[0m\n\u001b[0;32m    339\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39mproba_distribution\u001b[39m(\n\u001b[0;32m    340\u001b[0m     \u001b[39mself\u001b[39m: SelfMultiCategoricalDistribution, action_logits: th\u001b[39m.\u001b[39mTensor\n\u001b[0;32m    341\u001b[0m ) \u001b[39m-\u001b[39m\u001b[39m>\u001b[39m SelfMultiCategoricalDistribution:\n\u001b[1;32m--> 342\u001b[0m     \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mdistribution \u001b[39m=\u001b[39m [Categorical(logits\u001b[39m=\u001b[39msplit) \u001b[39mfor\u001b[39;00m split \u001b[39min\u001b[39;00m th\u001b[39m.\u001b[39msplit(action_logits, \u001b[39mtuple\u001b[39m(\u001b[39mself\u001b[39m\u001b[39m.\u001b[39maction_dims), dim\u001b[39m=\u001b[39m\u001b[39m1\u001b[39m)]\n\u001b[0;32m    343\u001b[0m     \u001b[39mreturn\u001b[39;00m \u001b[39mself\u001b[39m\n",
            "File \u001b[1;32md:\\Users\\amirh\\anaconda3\\lib\\site-packages\\stable_baselines3\\common\\distributions.py:342\u001b[0m, in \u001b[0;36m<listcomp>\u001b[1;34m(.0)\u001b[0m\n\u001b[0;32m    339\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39mproba_distribution\u001b[39m(\n\u001b[0;32m    340\u001b[0m     \u001b[39mself\u001b[39m: SelfMultiCategoricalDistribution, action_logits: th\u001b[39m.\u001b[39mTensor\n\u001b[0;32m    341\u001b[0m ) \u001b[39m-\u001b[39m\u001b[39m>\u001b[39m SelfMultiCategoricalDistribution:\n\u001b[1;32m--> 342\u001b[0m     \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mdistribution \u001b[39m=\u001b[39m [Categorical(logits\u001b[39m=\u001b[39;49msplit) \u001b[39mfor\u001b[39;00m split \u001b[39min\u001b[39;00m th\u001b[39m.\u001b[39msplit(action_logits, \u001b[39mtuple\u001b[39m(\u001b[39mself\u001b[39m\u001b[39m.\u001b[39maction_dims), dim\u001b[39m=\u001b[39m\u001b[39m1\u001b[39m)]\n\u001b[0;32m    343\u001b[0m     \u001b[39mreturn\u001b[39;00m \u001b[39mself\u001b[39m\n",
            "File \u001b[1;32md:\\Users\\amirh\\anaconda3\\lib\\site-packages\\torch\\distributions\\categorical.py:62\u001b[0m, in \u001b[0;36mCategorical.__init__\u001b[1;34m(self, probs, logits, validate_args)\u001b[0m\n\u001b[0;32m     60\u001b[0m         \u001b[39mraise\u001b[39;00m \u001b[39mValueError\u001b[39;00m(\u001b[39m\"\u001b[39m\u001b[39m`logits` parameter must be at least one-dimensional.\u001b[39m\u001b[39m\"\u001b[39m)\n\u001b[0;32m     61\u001b[0m     \u001b[39m# Normalize\u001b[39;00m\n\u001b[1;32m---> 62\u001b[0m     \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mlogits \u001b[39m=\u001b[39m logits \u001b[39m-\u001b[39m logits\u001b[39m.\u001b[39;49mlogsumexp(dim\u001b[39m=\u001b[39;49m\u001b[39m-\u001b[39;49m\u001b[39m1\u001b[39;49m, keepdim\u001b[39m=\u001b[39;49m\u001b[39mTrue\u001b[39;49;00m)\n\u001b[0;32m     63\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_param \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mprobs \u001b[39mif\u001b[39;00m probs \u001b[39mis\u001b[39;00m \u001b[39mnot\u001b[39;00m \u001b[39mNone\u001b[39;00m \u001b[39melse\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mlogits\n\u001b[0;32m     64\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_num_events \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_param\u001b[39m.\u001b[39msize()[\u001b[39m-\u001b[39m\u001b[39m1\u001b[39m]\n",
            "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
          ]
        }
      ],
      "source": [
        "from stable_baselines3 import DQN, PPO\n",
        "from stable_baselines3.common.monitor import Monitor\n",
        "from stable_baselines3.common.results_plotter import load_results, ts2xy\n",
        "from stable_baselines3.common.env_checker import check_env\n",
        "import os \n",
        "\n",
        "timesteps = 1e7\n",
        "log_dir = \"./tmp/\"\n",
        "\n",
        "check_env(env, warn=True)\n",
        "env = Monitor(env, log_dir+'monitor.csv', override_existing=False)\n",
        "\n",
        "model = PPO(\"MlpPolicy\", env, verbose=1, \n",
        "            tensorboard_log=\"logs\", \n",
        "            learning_rate=1e-3)\n",
        "\n",
        "# model = DQN(\"MlpPolicy\", env, verbose=1, \n",
        "#             tensorboard_log=\"logs\", \n",
        "#             learning_rate=1e-3, \n",
        "#             buffer_size=int(1e6),\n",
        "#             batch_size=1024,\n",
        "#             gamma=0.95,\n",
        "#             tau=0.05,\n",
        "#             policy_kwargs=dict(net_arch=[512, 512, 512]))\n",
        "\n",
        "callback = SaveOnBestTrainingRewardCallback(check_freq=2e4, log_dir=log_dir)\n",
        "\n",
        "# Train the agent\n",
        "# model = model.load('best_model.zip', env)\n",
        "model.learn(total_timesteps=timesteps, progress_bar=False) #callback=callback"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "model.actor"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "model.critic"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "model = model.load('./best_model', env)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "results = pd.read_csv('./monitor.csv').reset_index()\n",
        "results.columns = [results.iloc[0]]\n",
        "results.drop([0], inplace=True)\n",
        "returns = np.array(results['r'].astype(float)).ravel()\n",
        "\n",
        "window = 100\n",
        "rolling_mean = pd.Series(returns).rolling(window).mean()\n",
        "std = pd.Series(returns).rolling(window).std()\n",
        "\n",
        "plt.figure(figsize=(8,5))\n",
        "plt.plot(returns)\n",
        "plt.plot(rolling_mean, color='red', linewidth=2.5)\n",
        "plt.fill_between(range(len(rolling_mean)),rolling_mean-std, rolling_mean+std, color='violet', alpha=0.5)\n",
        "\n",
        "plt.xlabel('Episode')\n",
        "plt.ylabel('Return')\n",
        "plt.grid('on')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "results = pd.read_csv('./monitor.csv').reset_index()\n",
        "results.columns = [results.iloc[0]]\n",
        "results.drop([0], inplace=True)\n",
        "returns = np.array(results['l'].astype(float)).ravel()\n",
        "\n",
        "window = 100\n",
        "rolling_mean = pd.Series(returns).rolling(window).mean()\n",
        "std = pd.Series(returns).rolling(window).std()\n",
        "\n",
        "plt.figure(figsize=(8,5))\n",
        "plt.plot(returns)\n",
        "plt.plot(rolling_mean, color='red', linewidth=2.5)\n",
        "plt.fill_between(range(len(rolling_mean)),rolling_mean-std, rolling_mean+std, color='violet', alpha=0.5)\n",
        "\n",
        "plt.xlabel('Episode')\n",
        "plt.ylabel('Episode Length / Charge Duration (s)')\n",
        "plt.grid('on')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 676
        },
        "id": "X59u3VZ2KaZk",
        "outputId": "177053e0-a1a2-4e8d-f1bd-3d37f6515fa8"
      },
      "outputs": [],
      "source": [
        "vec_env = model.get_env()\n",
        "state = vec_env.reset()\n",
        "\n",
        "pack_res=[]\n",
        "rewards = []\n",
        "actions = []\n",
        "\n",
        "done = False\n",
        "pack_res.append(state.reshape(3,6))\n",
        "while not done:\n",
        "    action, _states = model.predict(state, deterministic=True)\n",
        "    next_state, reward, done, info = vec_env.step(action)\n",
        "    state = next_state\n",
        "    pack_res.append(next_state.reshape(3,6))\n",
        "    rewards.append(reward)\n",
        "    actions.append(action)\n",
        "\n",
        "pack_res = np.array(pack_res)\n",
        "plt.figure(figsize=(12,8))\n",
        "plt.subplot(2,2,1)\n",
        "plt.plot(pack_res[:-1,0,:])\n",
        "plt.ylabel('pack SOC')\n",
        "plt.grid('on')\n",
        "plt.subplot(2,2,2)\n",
        "plt.plot(pack_res[:-1,1,:])\n",
        "plt.ylabel('pack voltage')\n",
        "plt.grid('on')\n",
        "plt.subplot(2,2,3)\n",
        "plt.plot(pack_res[:-1,2,:])\n",
        "plt.ylabel('pack core temp')\n",
        "plt.grid('on')\n",
        "plt.subplot(2,2,4)\n",
        "plt.plot(np.array(actions)[:-1].reshape(-1,6+1)[:,:-1])\n",
        "plt.plot(np.array(actions)[:-1].reshape(-1,6+1)[:,-1], \"--\", color='red')\n",
        "plt.ylabel('current')\n",
        "plt.grid('on')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "np.array(actions)[:-1].shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "plt.plot(np.abs(pack_res[:-1,0,:] - np.mean(pack_res[:-1,0,:], axis=-1).reshape(-1,1)).sum(-1))\n",
        "plt.xlabel('time')\n",
        "plt.ylabel('Balance Cost')\n",
        "plt.grid('on')\n",
        "plt.ylim([0,0.2])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "plt.plot(pack_res[:-1,0,:].max(-1) - pack_res[:-1,0,:].min(-1))\n",
        "plt.xlabel('time')\n",
        "plt.ylabel('SOC dev')\n",
        "plt.grid('on')\n",
        "plt.ylim([0,0.2])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "np.abs(pack_res[:-1,0,:] - np.mean(pack_res[:-1,0,:], axis=-1).reshape(-1,1))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "np.mean(pack_res[:-1,0,:])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "pack_res[-2,0,:], np.std(pack_res[-2,0,:])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "model.save('./last_model.pth')"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "6oLYPqkn-gP8"
      },
      "source": [
        "# Tianshu"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qeAJGEMP-n3v",
        "outputId": "c7450abd-ec79-4abf-ce77-d6cfec762ed1"
      },
      "outputs": [],
      "source": [
        "pip install tianshou"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-EmT7PN-9ldL",
        "outputId": "a6a5c809-381d-4f56-e875-e9181841f91f"
      },
      "outputs": [],
      "source": [
        "import gymnasium as gym\n",
        "import torch, numpy as np, torch.nn as nn\n",
        "from torch.utils.tensorboard import SummaryWriter\n",
        "import tianshou as ts"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "GskHo6R7-6-X"
      },
      "outputs": [],
      "source": [
        "lr, epoch, batch_size = 1e-3, 10, 64\n",
        "train_num, test_num = 10, 100\n",
        "gamma, n_step, target_freq = 0.9, 3, 320\n",
        "buffer_size = 20000\n",
        "eps_train, eps_test = 0.1, 0.05\n",
        "step_per_epoch, step_per_collect = 10000, 10\n",
        "logger = ts.utils.TensorboardLogger(SummaryWriter('log/dqn'))  # TensorBoard is supported!\n",
        "# For other loggers: https://tianshou.readthedocs.io/en/master/tutorials/logger.html"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "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": "3.10.9"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}