diff --git a/scripts/highway_planning.ipynb b/scripts/highway_planning.ipynb
index 4fd9a87c0..b1722e1bb 100644
--- a/scripts/highway_planning.ipynb
+++ b/scripts/highway_planning.ipynb
@@ -1,100 +1,100 @@
 {
-  "cells": [
-    {
-      "cell_type": "markdown",
-      "metadata": {
-        "id": "QKWvMXWMBEJA"
-      },
-      "source": [
-        "# Behavioural Planning for Autonomous Highway Driving\n",
-        "\n",
-        "We plan a trajectory using the _Optimistic Planning for Deterministic systems_ ([OPD](https://hal.inria.fr/hal-00830182)) algorithm.\n"
-      ]
-    },
-    {
-      "cell_type": "code",
-      "execution_count": null,
-      "metadata": {
-        "id": "s-ghXis6A_md"
-      },
-      "outputs": [],
-      "source": [
-        "#@title Imports for env, agent, and visualisation.\n",
-        "# Environment\n",
-        "!pip install highway-env\n",
-        "import gymnasium as gym\n",
-        "import highway_env\n",
-        "\n",
-        "# Agent\n",
-        "!pip install git+https://github.com/eleurent/rl-agents#egg=rl-agents\n",
-        "from rl_agents.agents.common.factory import agent_factory\n",
-        "\n",
-        "# Visualisation\n",
-        "import sys\n",
-        "from tqdm.notebook import trange\n",
-        "!pip install moviepy -U\n",
-        "!pip install imageio_ffmpeg\n",
-        "!pip install pyvirtualdisplay\n",
-        "!apt-get install -y xvfb ffmpeg\n",
-        "!git clone https://github.com/Farama-Foundation/HighwayEnv.git\n",
-        "sys.path.insert(0, './highway-env/scripts/')\n",
-        "from utils import record_videos, show_videos\n"
-      ]
-    },
-    {
-      "cell_type": "code",
-      "execution_count": null,
-      "metadata": {
-        "id": "bgNDDWwqCj8l"
-      },
-      "outputs": [],
-      "source": [
-        "#@title Run an episode\n",
-        "\n",
-        "# Make environment\n",
-        "env = gym.make(\"highway-fast-v0\", render_mode=\"rgb_array\")\n",
-        "env = record_videos(env)\n",
-        "(obs, info), done = env.reset(), False\n",
-        "\n",
-        "# Make agent\n",
-        "agent_config = {\n",
-        "    \"__class__\": \"<class 'rl_agents.agents.tree_search.deterministic.DeterministicPlannerAgent'>\",\n",
-        "    \"env_preprocessors\": [{\"method\":\"simplify\"}],\n",
-        "    \"budget\": 50,\n",
-        "    \"gamma\": 0.7,\n",
-        "}\n",
-        "agent = agent_factory(env, agent_config)\n",
-        "\n",
-        "# Run episode\n",
-        "for step in trange(env.unwrapped.config[\"duration\"], desc=\"Running...\"):\n",
-        "    action = agent.act(obs)\n",
-        "    obs, reward, done, truncated, info = env.step(action)\n",
-        "    \n",
-        "env.close()\n",
-        "show_videos()"
-      ]
-    }
-  ],
-  "metadata": {
-    "colab": {
-      "name": "highway-planning.ipynb",
-      "provenance": []
-    },
-    "kernelspec": {
-      "display_name": "Python 3",
-      "language": "python",
-      "name": "python3"
-    },
-    "pycharm": {
-      "stem_cell": {
-        "cell_type": "raw",
-        "metadata": {
-          "collapsed": false
-        },
-        "source": []
-      }
-    }
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "QKWvMXWMBEJA"
+   },
+   "source": [
+    "# Behavioural Planning for Autonomous Highway Driving\n",
+    "\n",
+    "We plan a trajectory using the _Optimistic Planning for Deterministic systems_ ([OPD](https://hal.inria.fr/hal-00830182)) algorithm.\n"
+   ]
   },
-  "nbformat": 4,
-  "nbformat_minor": 0
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "id": "s-ghXis6A_md"
+   },
+   "source": [
+    "#@title Imports for env, agent, and visualisation.\n",
+    "# Environment\n",
+    "!pip install highway-env\n",
+    "import gymnasium as gym\n",
+    "import highway_env\n",
+    "\n",
+    "# Agent\n",
+    "!pip install git+https://github.com/eleurent/rl-agents#egg=rl-agents\n",
+    "from rl_agents.agents.common.factory import agent_factory\n",
+    "\n",
+    "# Visualisation\n",
+    "import sys\n",
+    "from tqdm.notebook import trange\n",
+    "!pip install moviepy -U\n",
+    "!pip install imageio_ffmpeg\n",
+    "!pip install pyvirtualdisplay\n",
+    "!apt-get install -y xvfb ffmpeg\n",
+    "!git clone https://github.com/Farama-Foundation/HighwayEnv.git\n",
+    "sys.path.insert(0, './highway-env/scripts/')\n",
+    "from utils import record_videos, show_videos\n"
+   ],
+   "outputs": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "id": "bgNDDWwqCj8l"
+   },
+   "source": [
+    "#@title Run an episode\n",
+    "\n",
+    "# Make environment\n",
+    "env = gym.make(\"highway-fast-v0\", render_mode=\"rgb_array\")\n",
+    "env = record_videos(env)\n",
+    "(obs, info), done = env.reset(), False\n",
+    "\n",
+    "# Make agent\n",
+    "agent_config = {\n",
+    "    \"__class__\": \"<class 'rl_agents.agents.tree_search.deterministic.DeterministicPlannerAgent'>\",\n",
+    "    \"env_preprocessors\": [{\"method\":\"simplify\"}],\n",
+    "    \"budget\": 50,\n",
+    "    \"gamma\": 0.7,\n",
+    "}\n",
+    "agent = agent_factory(env, agent_config)\n",
+    "\n",
+    "# Run episode\n",
+    "for step in trange(env.unwrapped.config[\"duration\"], desc=\"Running...\"):\n",
+    "    action = agent.act(obs)\n",
+    "    obs, reward, done, truncated, info = env.step(action)\n",
+    "    \n",
+    "env.close()\n",
+    "show_videos()"
+   ],
+   "outputs": []
+  }
+ ],
+ "metadata": {
+  "colab": {
+   "name": "highway-planning.ipynb",
+   "provenance": []
+  },
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "pycharm": {
+   "stem_cell": {
+    "cell_type": "raw",
+    "metadata": {
+     "collapsed": false
+    },
+    "source": []
+   }
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
 }
diff --git a/scripts/intersection_social_dqn.ipynb b/scripts/intersection_social_dqn.ipynb
index e29f32acb..bbfb2aaf2 100644
--- a/scripts/intersection_social_dqn.ipynb
+++ b/scripts/intersection_social_dqn.ipynb
@@ -1,179 +1,179 @@
 {
-  "cells": [
-    {
-      "cell_type": "markdown",
-      "metadata": {
-        "id": "sepDWoBqdRMK"
-      },
-      "source": [
-        "# Training a DQN with social attention on `intersection-v0`\n",
-        "\n"
-      ]
-    },
-    {
-      "cell_type": "code",
-      "execution_count": null,
-      "metadata": {
-        "id": "Kx8X4s8krNWt"
-      },
-      "outputs": [],
-      "source": [
-        "#@title Import requirements\n",
-        "\n",
-        "# Environment\n",
-        "!pip install highway-env\n",
-        "import gymnasium as gym\n",
-        "\n",
-        "# Agent\n",
-        "!pip install git+https://github.com/eleurent/rl-agents#egg=rl-agents\n",
-        "\n",
-        "# Visualisation utils\n",
-        "!pip install moviepy\n",
-        "!pip install imageio_ffmpeg\n",
-        "import sys\n",
-        "%load_ext tensorboard\n",
-        "!pip install tensorboardx gym pyvirtualdisplay\n",
-        "!apt-get install -y xvfb ffmpeg\n",
-        "!git clone https://github.com/Farama-Foundation/HighwayEnv.git 2> /dev/null\n",
-        "sys.path.insert(0, '/content/HighwayEnv/scripts/')\n",
-        "from utils import show_videos"
-      ]
-    },
-    {
-      "cell_type": "markdown",
-      "metadata": {
-        "id": "vvOEW00pdHrG"
-      },
-      "source": [
-        "## Training\n",
-        "\n",
-        "We use a policy architecture based on social attention, see [[Leurent and Mercat, 2019]](https://arxiv.org/abs/1911.12250).\n"
-      ]
-    },
-    {
-      "cell_type": "code",
-      "execution_count": null,
-      "metadata": {
-        "id": "QowKW3ix45ZW"
-      },
-      "outputs": [],
-      "source": [
-        "#@title Prepare environment, agent, and evaluation process.\n",
-        "\n",
-        "NUM_EPISODES = 3000  #@param {type: \"integer\"}\n",
-        "\n",
-        "from rl_agents.trainer.evaluation import Evaluation\n",
-        "from rl_agents.agents.common.factory import load_agent, load_environment\n",
-        "\n",
-        "# Get the environment and agent configurations from the rl-agents repository\n",
-        "!git clone https://github.com/eleurent/rl-agents.git 2> /dev/null\n",
-        "%cd /content/rl-agents/scripts/\n",
-        "env_config = 'configs/IntersectionEnv/env.json'\n",
-        "agent_config = 'configs/IntersectionEnv/agents/DQNAgent/ego_attention_2h.json'\n",
-        "\n",
-        "env = load_environment(env_config)\n",
-        "agent = load_agent(agent_config, env)\n",
-        "evaluation = Evaluation(env, agent, num_episodes=NUM_EPISODES, display_env=False, display_agent=False)\n",
-        "print(f\"Ready to train {agent} on {env}\")"
-      ]
-    },
-    {
-      "cell_type": "markdown",
-      "metadata": {
-        "id": "nqnGqW6jd1xN"
-      },
-      "source": [
-        "Run tensorboard locally to visualize training."
-      ]
-    },
-    {
-      "cell_type": "code",
-      "execution_count": null,
-      "metadata": {
-        "id": "q7QJY2wc4_1N"
-      },
-      "outputs": [],
-      "source": [
-        "%tensorboard --logdir \"{evaluation.directory}\""
-      ]
-    },
-    {
-      "cell_type": "markdown",
-      "metadata": {
-        "id": "BtK9dtfb0JMF"
-      },
-      "source": [
-        "Start training. This should take about an hour."
-      ]
-    },
-    {
-      "cell_type": "code",
-      "execution_count": null,
-      "metadata": {
-        "id": "sFVq1gFz42Eg"
-      },
-      "outputs": [],
-      "source": [
-        "evaluation.train()"
-      ]
-    },
-    {
-      "cell_type": "markdown",
-      "metadata": {
-        "id": "-lNvWg42RWiw"
-      },
-      "source": [
-        "Progress can be visualised in the tensorboard cell above, which should update every 30s (or manually). You may need to click the *Fit domain to data* buttons below each graph."
-      ]
-    },
-    {
-      "cell_type": "markdown",
-      "metadata": {
-        "id": "VKfvu5uhzCIU"
-      },
-      "source": [
-        "## Testing"
-      ]
-    },
-    {
-      "cell_type": "code",
-      "execution_count": null,
-      "metadata": {
-        "id": "gY0rpVYUtRpN"
-      },
-      "outputs": [],
-      "source": [
-        "#@title Run the learned policy for a few episodes.\n",
-        "env = load_environment(env_config)\n",
-        "env.config[\"offscreen_rendering\"] = True\n",
-        "agent = load_agent(agent_config, env)\n",
-        "evaluation = Evaluation(env, agent, num_episodes=20, training = False, recover = True)\n",
-        "evaluation.test()\n",
-        "show_videos(evaluation.run_directory)"
-      ]
-    }
-  ],
-  "metadata": {
-    "accelerator": "GPU",
-    "colab": {
-      "name": "SocialAttentionDQN",
-      "provenance": []
-    },
-    "kernelspec": {
-      "display_name": "Python 3",
-      "language": "python",
-      "name": "python3"
-    },
-    "pycharm": {
-      "stem_cell": {
-        "cell_type": "raw",
-        "metadata": {
-          "collapsed": false
-        },
-        "source": []
-      }
-    }
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "sepDWoBqdRMK"
+   },
+   "source": [
+    "# Training a DQN with social attention on `intersection-v0`\n",
+    "\n"
+   ]
   },
-  "nbformat": 4,
-  "nbformat_minor": 0
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "id": "Kx8X4s8krNWt"
+   },
+   "source": [
+    "#@title Import requirements\n",
+    "\n",
+    "# Environment\n",
+    "!pip install highway-env\n",
+    "import gymnasium as gym\n",
+    "\n",
+    "# Agent\n",
+    "!pip install git+https://github.com/eleurent/rl-agents#egg=rl-agents\n",
+    "\n",
+    "# Visualisation utils\n",
+    "!pip install moviepy\n",
+    "!pip install imageio_ffmpeg\n",
+    "import sys\n",
+    "%load_ext tensorboard\n",
+    "!pip install tensorboardx gym pyvirtualdisplay\n",
+    "!apt-get install -y xvfb ffmpeg\n",
+    "!git clone https://github.com/Farama-Foundation/HighwayEnv.git 2> /dev/null\n",
+    "sys.path.insert(0, '/content/HighwayEnv/scripts/')\n",
+    "from utils import show_videos"
+   ],
+   "outputs": []
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "vvOEW00pdHrG"
+   },
+   "source": [
+    "## Training\n",
+    "\n",
+    "We use a policy architecture based on social attention, see [[Leurent and Mercat, 2019]](https://arxiv.org/abs/1911.12250).\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "id": "QowKW3ix45ZW"
+   },
+   "source": [
+    "#@title Prepare environment, agent, and evaluation process.\n",
+    "\n",
+    "NUM_EPISODES = 3000  #@param {type: \"integer\"}\n",
+    "\n",
+    "from rl_agents.trainer.evaluation import Evaluation\n",
+    "from rl_agents.agents.common.factory import load_agent, load_environment\n",
+    "\n",
+    "# Get the environment and agent configurations from the rl-agents repository\n",
+    "!git clone https://github.com/eleurent/rl-agents.git 2> /dev/null\n",
+    "%cd /content/rl-agents/scripts/\n",
+    "env_config = 'configs/IntersectionEnv/env.json'\n",
+    "agent_config = 'configs/IntersectionEnv/agents/DQNAgent/ego_attention_2h.json'\n",
+    "\n",
+    "env = load_environment(env_config)\n",
+    "agent = load_agent(agent_config, env)\n",
+    "evaluation = Evaluation(env, agent, num_episodes=NUM_EPISODES, display_env=False, display_agent=False)\n",
+    "print(f\"Ready to train {agent} on {env}\")"
+   ],
+   "outputs": []
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "nqnGqW6jd1xN"
+   },
+   "source": [
+    "Run tensorboard locally to visualize training."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "id": "q7QJY2wc4_1N"
+   },
+   "source": [
+    "%tensorboard --logdir \"{evaluation.directory}\""
+   ],
+   "outputs": []
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "BtK9dtfb0JMF"
+   },
+   "source": [
+    "Start training. This should take about an hour."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "id": "sFVq1gFz42Eg"
+   },
+   "source": [
+    "evaluation.train()"
+   ],
+   "outputs": []
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "-lNvWg42RWiw"
+   },
+   "source": [
+    "Progress can be visualised in the tensorboard cell above, which should update every 30s (or manually). You may need to click the *Fit domain to data* buttons below each graph."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "VKfvu5uhzCIU"
+   },
+   "source": [
+    "## Testing"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "id": "gY0rpVYUtRpN"
+   },
+   "source": [
+    "#@title Run the learned policy for a few episodes.\n",
+    "env = load_environment(env_config)\n",
+    "env.config[\"offscreen_rendering\"] = True\n",
+    "agent = load_agent(agent_config, env)\n",
+    "evaluation = Evaluation(env, agent, num_episodes=20, training = False, recover = True)\n",
+    "evaluation.test()\n",
+    "show_videos(evaluation.run_directory)"
+   ],
+   "outputs": []
+  }
+ ],
+ "metadata": {
+  "accelerator": "GPU",
+  "colab": {
+   "name": "SocialAttentionDQN",
+   "provenance": []
+  },
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "pycharm": {
+   "stem_cell": {
+    "cell_type": "raw",
+    "metadata": {
+     "collapsed": false
+    },
+    "source": []
+   }
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
 }