Deep Q Learning.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Deep Q Learning Algorithm Implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import random\n",
    "import gym\n",
    "\n",
    "import numpy as np\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from torchvision import transforms\n",
    "from collections import deque\n",
    "\n",
    "from IPython.display import clear_output\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "LEARNING_RATE = 0.0001\n",
    "DISCOUNT = 0.99\n",
    "EPISODES = 25000\n",
    "SHOW_EVERY = 300\n",
    "BATCH_SIZE = 4\n",
    "REWARD_PTH = 'rewards.npy'\n",
    "ENV_NAME ='CartPole-v0'\n",
    "MODEL_PTH = './models/' + ENV_NAME + '.pth'\n",
    "\n",
    "STARTING_EPSILON = 0.9\n",
    "END_EPSILON = 0.1\n",
    "DECAY_RATE = 0.9999\n",
    "BATCH_SIZE = 64\n",
    "\n",
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# One hidden layer function approximator\n",
    "class Actor(nn.Module):\n",
    "\n",
    "    def __init__(self, starting_epsilon, end_epsilon, decay_rate, batch_size):\n",
    "        super(Actor, self).__init__()\n",
    "        self. eps = starting_epsilon\n",
    "        self.eps_end = end_epsilon\n",
    "        self.eps_decay = decay_rate\n",
    "        self.criterion = torch.nn.MSELoss()\n",
    "        self.batch_size = batch_size       \n",
    "        self.block = nn.Sequential(\n",
    "            nn.Linear(4, 64),\n",
    "            nn.ReLU(),\n",
    "            nn.Linear(64, 32),\n",
    "            nn.ReLU(),\n",
    "            nn.Linear(32, 2),\n",
    "        )\n",
    "\n",
    "    def forward(self, ip):\n",
    "        return self.block(ip)\n",
    "    \n",
    "    \n",
    "    def act(self, ip):\n",
    "        self.eps = max(self.eps*self.eps_decay, self.eps_end)      # Decay Doubt Batch random acting doubt\n",
    "        ip   = torch.FloatTensor(ip).unsqueeze(0).to(device)\n",
    "\n",
    "        if random.random() < self.eps:\n",
    "            return env.action_space.sample()\n",
    "        else:\n",
    "            return self.forward(ip).max(dim=1)[1].item()\n",
    "        \n",
    "    def update(self, memory, optimizer, target_net, loss):\n",
    "        \n",
    "        if len(memory) < self.batch_size:\n",
    "            return 0 \n",
    "        \n",
    "        states, actions, rewards, next_states, dones = memory.sample(self.batch_size)\n",
    "        \n",
    "        states      = (torch.FloatTensor(np.float32(states))).to(device)\n",
    "        next_states = (torch.FloatTensor(np.float32(next_states))).to(device)\n",
    "        actions     = (torch.LongTensor(actions)).unsqueeze(1).to(device)\n",
    "        rewards     = (torch.FloatTensor(rewards)).to(device)\n",
    "        dones       = (torch.FloatTensor(dones)).to(device)\n",
    "\n",
    "        q = self.forward(states).gather(1, actions).squeeze(1)\n",
    "        q_expected = rewards + target_net(next_states).max(dim=1)[0] * (1 - dones)\n",
    "        \n",
    "        td_loss = self.criterion(q, q_expected.detach())\n",
    "        loss += td_loss.item()\n",
    "        \n",
    "        optimizer.zero_grad()\n",
    "        td_loss.backward()\n",
    "        optimizer.step()\n",
    "        \n",
    "        return loss\n",
    "        \n",
    "    \n",
    "# def weights_init(m):\n",
    "#     classname = m.__class__.__name__\n",
    "#     if classname.find('Linear') != -1:\n",
    "#         nn.init.normal_(m.weight, 0, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# One hidden layer function approximator\n",
    "class Target(nn.Module):\n",
    "\n",
    "    def __init__(self, starting_epsilon, end_epsilon, decay_rate, batch_size):\n",
    "        super(Target, self).__init__()\n",
    "        self. eps = starting_epsilon\n",
    "        self.eps_end = end_epsilon\n",
    "        self.eps_decay = decay_rate\n",
    "        self.criterion = torch.nn.MSELoss()\n",
    "        self.batch_size = batch_size       \n",
    "        self.block = nn.Sequential(\n",
    "            nn.Linear(4, 64),\n",
    "            nn.ReLU(),\n",
    "            nn.Linear(64, 32),\n",
    "            nn.ReLU(),\n",
    "            nn.Linear(32, 2),\n",
    "        )\n",
    "\n",
    "    def forward(self, ip):\n",
    "        return self.block(ip)\n",
    "    \n",
    "    \n",
    "    def act(self, ip):\n",
    "        self.eps = max(self.eps*self.eps_decay, self.eps_end)      # Decay Doubt Batch random acting doubt\n",
    "        ip   = torch.FloatTensor(ip).unsqueeze(0).to(device)\n",
    "\n",
    "        if random.random() < self.eps:\n",
    "            return env.action_space.sample()\n",
    "        else:\n",
    "            return self.forward(ip).max(dim=1)[1].item()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ReplayBuffer(object):\n",
    "    def __init__(self, capacity):\n",
    "        self.buffer = deque(maxlen=capacity)\n",
    "    \n",
    "    def push(self, state, action, reward, next_state, done):\n",
    "        state      = np.expand_dims(state, 0)\n",
    "        next_state = np.expand_dims(next_state, 0)\n",
    "            \n",
    "        self.buffer.append((state, action, reward, next_state, done))\n",
    "    \n",
    "    def sample(self, batch_size):\n",
    "        state, action, reward, next_state, done = zip(*random.sample(self.buffer, batch_size))\n",
    "        return np.concatenate(state), action, reward, np.concatenate(next_state), done\n",
    "    \n",
    "    def __len__(self):\n",
    "        return len(self.buffer)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Observations: [-0.04454075  0.03747024  0.00458636 -0.04870187]\n",
      "Observations Limit: [-4.8000002e+00 -3.4028235e+38 -4.1887903e-01 -3.4028235e+38] to [4.8000002e+00 3.4028235e+38 4.1887903e-01 3.4028235e+38] | Number of action values: 2\n"
     ]
    }
   ],
   "source": [
    "# Instaciate action_value_function class\n",
    "q_hat = Actor(STARTING_EPSILON, END_EPSILON, DECAY_RATE, BATCH_SIZE)\n",
    "q_hat_target = Target(STARTING_EPSILON, END_EPSILON, DECAY_RATE, BATCH_SIZE).to(device)\n",
    "q_hat_target.load_state_dict(q_hat.state_dict())\n",
    "q_hat.to(device)\n",
    "q_hat_target.to(device)\n",
    "\n",
    "memory = ReplayBuffer(10000)\n",
    "\n",
    "env = gym.make(ENV_NAME)\n",
    "\n",
    "env.reset()\n",
    "print('Observations:', env.reset())\n",
    "print('Observations Limit:',env.observation_space.low,'to',env.observation_space.high, '| Number of action values:',env.action_space.n)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "optimizer = optim.Adam(q_hat.parameters())\n",
    "# scheduler = optim.lr_scheduler.StepLR(optimizer, step_size=1, gamma=0.9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot(episode, avg_loss, eps, eps_rewards):\n",
    "    clear_output(True)\n",
    "    plt.figure(figsize=(20,5))\n",
    "    plt.subplot(131)\n",
    "    plt.title('Episode: %5d | Average Loss: %5.2f | Epsilon: %4.2f | Avg. Reward: %5.2f'%(episode, np.mean(ep_loss[-20:]), eps, np.mean(ep_rewards[-20:])))\n",
    "    plt.plot(eps_rewards)\n",
    "    plt.subplot(132)\n",
    "    plt.title('loss')\n",
    "    plt.plot(ep_loss)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-20-38b385e6ea16>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m     15\u001b[0m     \u001b[0;32mwhile\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mdone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     16\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 17\u001b[0;31m         \u001b[0maction\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mq_hat\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mact\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstate\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     18\u001b[0m         \u001b[0mnext_state\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreward\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0m_\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0menv\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0maction\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     19\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-3-87729fe1f690>\u001b[0m in \u001b[0;36mact\u001b[0;34m(self, ip)\u001b[0m\n\u001b[1;32m     28\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0menv\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0maction_space\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msample\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     29\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 30\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mip\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmax\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdim\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     31\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     32\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mupdate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmemory\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtarget_net\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mloss\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-3-87729fe1f690>\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, ip)\u001b[0m\n\u001b[1;32m     18\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     19\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mip\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 20\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mblock\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mip\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     21\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     22\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/projects/venv/lib/python3.6/site-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m    539\u001b[0m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    540\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 541\u001b[0;31m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    542\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mhook\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    543\u001b[0m             \u001b[0mhook_result\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mhook\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mresult\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/projects/venv/lib/python3.6/site-packages/torch/nn/modules/container.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, input)\u001b[0m\n\u001b[1;32m     90\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     91\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mmodule\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_modules\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 92\u001b[0;31m             \u001b[0minput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodule\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     93\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     94\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/projects/venv/lib/python3.6/site-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m    540\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    541\u001b[0m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 542\u001b[0;31m         \u001b[0;32mfor\u001b[0m \u001b[0mhook\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    543\u001b[0m             \u001b[0mhook_result\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mhook\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mresult\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    544\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mhook_result\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "avg_loss = 0\n",
    "STEPS = 200\n",
    "ep_rewards = np.array([])\n",
    "ep_loss = np.array([])\n",
    "\n",
    "# Train agent\n",
    "for episode in range(EPISODES):\n",
    "    # Episode wise inits\n",
    "    state = env.reset()\n",
    "    ep_reward = 0\n",
    "    running_loss = 0\n",
    "    done = 0\n",
    "    \n",
    "    # Collect Trajectory\n",
    "    while not done:\n",
    "        \n",
    "        action = q_hat.act(state)\n",
    "        next_state, reward, done, _ = env.step(action)\n",
    "                    \n",
    "        ep_reward += reward\n",
    "        memory.push(state, action, reward, next_state, done)\n",
    "       \n",
    "        state = next_state\n",
    "    \n",
    "        #Update\n",
    "        running_loss = q_hat.update(memory, optimizer, q_hat_target, running_loss)\n",
    "    \n",
    "    ep_rewards = np.append(ep_rewards, ep_reward)\n",
    "    ep_loss = np.append(ep_loss, running_loss)\n",
    "    q_hat_target.load_state_dict(q_hat.state_dict())\n",
    "\n",
    "    if episode%20 == 0:\n",
    "        plot(episode, ep_loss, q_hat.eps, ep_rewards)\n",
    "        avg_loss = 0\n",
    "        ep_reward = 0\n",
    "        np.save(REWARD_PTH, ep_rewards)\n",
    "\n",
    "env.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save model\n",
    "torch.save(q_hat_target.state_dict(),MODEL_PTH)\n",
    "import torch\n",
    "q_hat_target = Target(STARTING_EPSILON, END_EPSILON, DECAY_RATE, BATCH_SIZE)\n",
    "q_hat_target.load_state_dict(torch.load(MODEL_PTH))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Reward: 200.0\n"
     ]
    }
   ],
   "source": [
    "# Test your policy\n",
    "import gym\n",
    "import time\n",
    "\n",
    "env = gym.make(ENV_NAME)\n",
    "# env.reset()\n",
    "state = env.reset()\n",
    "done = False\n",
    "EPS_TEST = 0.05\n",
    "tot_reward = 0\n",
    "\n",
    "for step in range(2000):\n",
    "    state = torch.from_numpy(state).float().to(device)\n",
    "    env.render()\n",
    "    action = torch.argmax(q_hat_target(state)).item()\n",
    "    state, reward, done, _ = env.step(action)\n",
    "    tot_reward += reward\n",
    "    if done:\n",
    "#         print('DONE')\n",
    "        break\n",
    "    time.sleep(.002)\n",
    "    \n",
    "print('Total Reward:', tot_reward)\n",
    "    \n",
    "env.close()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}