RL_CarRacing.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "import gym\n",
    "from stable_baselines3 import PPO\n",
    "from stable_baselines3.common.vec_env import DummyVecEnv\n",
    "from stable_baselines3.common.evaluation import evaluate_policy\n",
    "import os"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Test environment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "environment_name = 'CarRacing-v0'\n",
    "env = gym.make(environment_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Track generation: 1033..1300 -> 267-tiles track\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[[0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0]],\n",
       "\n",
       "       [[0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0]],\n",
       "\n",
       "       [[0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0]],\n",
       "\n",
       "       ...,\n",
       "\n",
       "       [[0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0]],\n",
       "\n",
       "       [[0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0]],\n",
       "\n",
       "       [[0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0],\n",
       "        [0, 0, 0]]], dtype=uint8)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "env.reset()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Box([-1.  0.  0.], [1. 1. 1.], (3,), float32)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "env.action_space"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Box([[[0 0 0]\n",
       "  [0 0 0]\n",
       "  [0 0 0]\n",
       "  ...\n",
       "  [0 0 0]\n",
       "  [0 0 0]\n",
       "  [0 0 0]]\n",
       "\n",
       " [[0 0 0]\n",
       "  [0 0 0]\n",
       "  [0 0 0]\n",
       "  ...\n",
       "  [0 0 0]\n",
       "  [0 0 0]\n",
       "  [0 0 0]]\n",
       "\n",
       " [[0 0 0]\n",
       "  [0 0 0]\n",
       "  [0 0 0]\n",
       "  ...\n",
       "  [0 0 0]\n",
       "  [0 0 0]\n",
       "  [0 0 0]]\n",
       "\n",
       " ...\n",
       "\n",
       " [[0 0 0]\n",
       "  [0 0 0]\n",
       "  [0 0 0]\n",
       "  ...\n",
       "  [0 0 0]\n",
       "  [0 0 0]\n",
       "  [0 0 0]]\n",
       "\n",
       " [[0 0 0]\n",
       "  [0 0 0]\n",
       "  [0 0 0]\n",
       "  ...\n",
       "  [0 0 0]\n",
       "  [0 0 0]\n",
       "  [0 0 0]]\n",
       "\n",
       " [[0 0 0]\n",
       "  [0 0 0]\n",
       "  [0 0 0]\n",
       "  ...\n",
       "  [0 0 0]\n",
       "  [0 0 0]\n",
       "  [0 0 0]]], [[[255 255 255]\n",
       "  [255 255 255]\n",
       "  [255 255 255]\n",
       "  ...\n",
       "  [255 255 255]\n",
       "  [255 255 255]\n",
       "  [255 255 255]]\n",
       "\n",
       " [[255 255 255]\n",
       "  [255 255 255]\n",
       "  [255 255 255]\n",
       "  ...\n",
       "  [255 255 255]\n",
       "  [255 255 255]\n",
       "  [255 255 255]]\n",
       "\n",
       " [[255 255 255]\n",
       "  [255 255 255]\n",
       "  [255 255 255]\n",
       "  ...\n",
       "  [255 255 255]\n",
       "  [255 255 255]\n",
       "  [255 255 255]]\n",
       "\n",
       " ...\n",
       "\n",
       " [[255 255 255]\n",
       "  [255 255 255]\n",
       "  [255 255 255]\n",
       "  ...\n",
       "  [255 255 255]\n",
       "  [255 255 255]\n",
       "  [255 255 255]]\n",
       "\n",
       " [[255 255 255]\n",
       "  [255 255 255]\n",
       "  [255 255 255]\n",
       "  ...\n",
       "  [255 255 255]\n",
       "  [255 255 255]\n",
       "  [255 255 255]]\n",
       "\n",
       " [[255 255 255]\n",
       "  [255 255 255]\n",
       "  [255 255 255]\n",
       "  ...\n",
       "  [255 255 255]\n",
       "  [255 255 255]\n",
       "  [255 255 255]]], (96, 96, 3), uint8)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "env.observation_space"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "env.render()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "env.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "episodes = 5 #Try the environment 5 times\n",
    "for episode in range(1, episodes+1):\n",
    "    state = env.reset() #Initial set of observations\n",
    "    done = False\n",
    "    score = 0\n",
    "\n",
    "    while not done:\n",
    "        env.render() #View the graphical representation of env\n",
    "        action = env.action_space.sample() #Random selection of an action\n",
    "        n_state, reward, done, info = env.step(action)\n",
    "        score += reward\n",
    "    print('Episode:{} Score:{}'.format(episode, score))\n",
    "env.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "env = gym.make(environment_name)\n",
    "env = DummyVecEnv([lambda: env])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using cpu device\n",
      "Wrapping the env in a VecTransposeImage.\n"
     ]
    }
   ],
   "source": [
    "log_path = os.path.join('Training', 'Logs')\n",
    "model = PPO('CnnPolicy', env, verbose=1, tensorboard_log=log_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.learn(total_timesteps=100000)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "ppo_path = os.path.join('Training', 'Saved_Models', 'PPO_Driving_model')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save(ppo_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "del model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wrapping the env in a VecTransposeImage.\n"
     ]
    }
   ],
   "source": [
    "model = PPO.load(ppo_path, env)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluate and test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "evaluate_policy(model, env, n_eval_episodes=10, render=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "env.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Test model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "episodes = 5 #Try the environment 5 times\n",
    "for episode in range(1, episodes+1):\n",
    "    obs = env.reset() \n",
    "    done = False\n",
    "    score = 0\n",
    "\n",
    "    while not done:\n",
    "        env.render() \n",
    "        action, _ = model.predict(obs) # USING MODEL HERE\n",
    "        obs, reward, done, info = env.step(action)\n",
    "        score += reward\n",
    "    print('Episode:{} Score:{}'.format(episode, score))\n",
    "env.close()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}