InvPEN.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "InvPEN.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "z6eAE_SoTrUV"
      },
      "outputs": [],
      "source": [
        "import gym\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "import random\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "from torch.distributions.normal import Normal\n",
        "from torch.autograd import Variable"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "class Actor_Network(nn.Module):\n",
        "    \"\"\"\n",
        "    Gaussian Policy\n",
        "    Outputs means and std deviation of a gaussian policy\n",
        "    \"\"\"\n",
        "    def __init__(self, input_layer, output_layer):\n",
        "        super(Actor_Network,self).__init__()\n",
        "        self.linear1 = nn.Linear(input_layer,32)\n",
        "        self.linear2 = nn.Linear(32,128)\n",
        "        self.linear3 = nn.Linear(128,32)\n",
        "        self.linear4 = nn.Linear(32,output_layer)\n",
        "\n",
        "    def forward(self,x):\n",
        "        x = F.relu(self.linear1(x))\n",
        "        x = F.relu(self.linear2(x))\n",
        "        x = F.relu(self.linear3(x))\n",
        "        #x = F.relu(self.linear4(x))\n",
        "        linear_out = self.linear4(x)\n",
        "        #s = nn.Softplus()\n",
        "        #x = s(linear_out)\n",
        "        #print(linear_out)\n",
        "        linear_out[1] = F.relu(linear_out[1])+0.00001\n",
        "        #print(linear_out)\n",
        "\n",
        "        return linear_out\n",
        "        #return x\n",
        "\n",
        "class Critic_Network(nn.Module):\n",
        "    \"\"\"\n",
        "    Outputs Value Function\n",
        "    \"\"\"\n",
        "    def __init__(self,input_layer):\n",
        "        super(Critic_Network,self).__init__()\n",
        "        self.linear1 = nn.Linear(input_layer,32)\n",
        "        self.linear2 = nn.Linear(32,128)\n",
        "        self.linear3 = nn.Linear(128,32)\n",
        "        self.linear4 = nn.Linear(32,1)\n",
        "        \n",
        "\n",
        "    def forward(self,x):\n",
        "        x = F.relu(self.linear1(x))\n",
        "        x = F.relu(self.linear2(x))\n",
        "        x = F.relu(self.linear3(x))\n",
        "        x = self.linear4(x)\n",
        "        return x"
      ],
      "metadata": {
        "id": "0yK3wmiVTv0V"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "class ReplayMemory():\n",
        "    def __init__(self,size):\n",
        "        self.size = size\n",
        "        #state, action, rew, next_state, done\n",
        "        self.memory = [[],[],[],[],[]]\n",
        "    def store(self,data):\n",
        "        if len(self.memory[0])==self.size:\n",
        "            for idx in range(5):\n",
        "                self.memory[idx].pop(0)\n",
        "        for idx,part in enumerate(data):\n",
        "            self.memory[idx].append(part)\n",
        "    def sample(self,batch_size):\n",
        "        rows = random.sample(range(0,len(self.memory[0])),batch_size)\n",
        "        experiences = [[],[],[],[],[]]\n",
        "        for row in rows:\n",
        "            for col in range(5):\n",
        "                experiences[col].append(self.memory[col][row])\n",
        "        return experiences\n",
        "\n",
        "    def __len__(self):\n",
        "        return len(self.memory[0])"
      ],
      "metadata": {
        "id": "28oyW-mDTwlU"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "if(torch.cuda.is_available()):\n",
        "    print('CUDA Available')\n",
        "device = torch.device('cuda')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lsgUYOT0UZiA",
        "outputId": "c50c2129-7835-4d16-ad2e-f6f337b490c4"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "CUDA Available\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#actor = Actor_Network(2,2).to(device)\n",
        "\n",
        "#print(actor(torch.FloatTensor([5,6])), actor(torch.FloatTensor([5,6])).detach().numpy().squeeze())\n",
        "\n",
        "\n",
        "env = gym.make('Pendulum-v0')\n",
        "actor_model = Actor_Network(3,2).to(device)\n",
        "critic_model = Critic_Network(3).to(device)\n",
        "critic_target_model = Critic_Network(3).to(device)\n",
        "critic_target_model.eval()\n",
        "\n",
        "gamma = 0.9\n",
        "learning_rate = 0.0005\n",
        "critic_optimizer = torch.optim.Adam(critic_model.parameters(),lr=learning_rate)\n",
        "actor_optimizer = torch.optim.Adam(actor_model.parameters(),lr=learning_rate)\n",
        "\n",
        "memory = ReplayMemory(5000)\n",
        "BATCH_SIZE = 64"
      ],
      "metadata": {
        "id": "C8TPijjKUHvN"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def train_critic():\n",
        "    #print(len(memory))\n",
        "    if len(memory)<BATCH_SIZE:\n",
        "        return 0\n",
        "    experiences = memory.sample(BATCH_SIZE)\n",
        "    \n",
        "    state_batch = np.array(experiences[0])\n",
        "    state_batch = torch.Tensor(state_batch).to(device)\n",
        "\n",
        "    #action_batch = torch.Tensor(experiences[1])  # .unsqueeze(1)\n",
        "    reward_batch = np.array(experiences[2])\n",
        "    reward_batch = torch.Tensor(reward_batch)\n",
        "\n",
        "    next_state_batch = np.array(experiences[3])\n",
        "    next_state_batch = torch.Tensor(next_state_batch).to(device)\n",
        "    done_batch = experiences[4]\n",
        "\n",
        "    predicted_Vs = critic_model(state_batch)\n",
        "    predicted_Vs = predicted_Vs.reshape(BATCH_SIZE)\n",
        "    with torch.no_grad():\n",
        "        target_Vns = critic_target_model(next_state_batch.to(device)).cpu()\n",
        "\n",
        "    target_Vs = torch.zeros(BATCH_SIZE).to(device)\n",
        "    for i,is_done in enumerate(done_batch):\n",
        "        if is_done:\n",
        "            target_Vs[i] = reward_batch[i]\n",
        "        else:\n",
        "            target_Vs[i] = reward_batch[i] + gamma*target_Vns[i]\n",
        "    \n",
        "    critic_optimizer.zero_grad()\n",
        "    loss = nn.MSELoss()\n",
        "    critic_loss = loss(predicted_Vs,target_Vs)\n",
        "    #print(critic_loss)\n",
        "    #print('predicted_Vs',predicted_Vs)\n",
        "    #print('target_Vs',target_Vs)\n",
        "\n",
        "    critic_loss.backward()\n",
        "    critic_optimizer.step()\n",
        "    return critic_loss"
      ],
      "metadata": {
        "id": "9TNQpqHkUJr9"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def train_actor(step_data):\n",
        "    state = step_data[0]\n",
        "    action = step_data[1]\n",
        "    rew = step_data[2]\n",
        "    new_state = step_data[3]\n",
        "\n",
        "    with torch.no_grad():\n",
        "        Vs = critic_model(torch.FloatTensor(state).to(device))\n",
        "        Vns = critic_model(torch.FloatTensor(new_state).to(device))\n",
        "    act_val = torch.from_numpy(np.array(rew)).float().to(device) + gamma*Vns - Vs\n",
        "\n",
        "    dist_param = actor_model(torch.FloatTensor(state).to(device))\n",
        "    #if dist_param[1] <= 0:\n",
        "    #    dev = 0.0001\n",
        "    #else:\n",
        "    #    dev = dist_param[1]\n",
        "    dist = Normal(dist_param[0],dist_param[1])\n",
        "\n",
        "    actor_optimizer.zero_grad()\n",
        "    actor_loss = -dist.log_prob(torch.from_numpy(action).float().to(device))*act_val\n",
        "\n",
        "    actor_loss.backward()\n",
        "    actor_optimizer.step()\n",
        "    return actor_loss"
      ],
      "metadata": {
        "id": "vlBCbxe_UNRA"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "change_model_after = 75\n",
        "rew_list = []\n",
        "critic_loss_list = []\n",
        "actor_loss_list = []\n",
        "for ep in range(1200):\n",
        "    state = env.reset()\n",
        "    if ep%change_model_after==0:\n",
        "        print('model_changed..')\n",
        "        critic_target_model.load_state_dict(critic_model.state_dict())\n",
        "    #print('state', state)\n",
        "    done = False\n",
        "    total_rew = 0\n",
        "    while not done:\n",
        "\n",
        "        with torch.no_grad():\n",
        "            dist_param = actor_model(torch.from_numpy(state).float().to(device))\n",
        "           #if dist_param[1] <= 0:\n",
        "           #     dist_param[1] = 0.0001\n",
        "            dist = Normal(dist_param[0], dist_param[1])\n",
        "            # action = dist.sample().item()\n",
        "            action = dist.sample().cpu().numpy().reshape(1)\n",
        "            apply_action = 2*action\n",
        "            # print(state)\n",
        "            # print(dist_param)\n",
        "        new_state, rew, done, info = env.step(apply_action)\n",
        "\n",
        "        total_rew += rew\n",
        "\n",
        "        memory.store([state,action,rew,new_state,done])\n",
        "        critic_loss = train_critic()\n",
        "        actor_loss = train_actor([state,action,rew,new_state])\n",
        "        critic_loss_list.append(critic_loss)\n",
        "        #print(critic_loss)\n",
        "        actor_loss_list.append(actor_loss)\n",
        "    print(ep,total_rew)\n",
        "    rew_list.append(total_rew)\n",
        "    #print(critic_loss,actor_loss)\n",
        "\n",
        "\n",
        "fig1,ax1 = plt.subplots()\n",
        "x = np.arange(len(rew_list))\n",
        "ax1.plot(x,rew_list)\n",
        "ax1.set_xlabel('episodes')\n",
        "ax1.set_ylabel('rewards')\n",
        "fig1loss,ax1loss = plt.subplots()\n",
        "x_loss = np.arange(len(critic_loss_list))\n",
        "critic_loss_list.cpu()\n",
        "ax1loss.plot(x_loss,critic_loss_list)\n",
        "ax1loss.set_xlabel('steps')\n",
        "ax1loss.set_ylabel('loss')\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "BGAypHXTUNol",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "cf8385b6-58ec-4682-d4fd-26b88f9a0f0b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "model_changed..\n",
            "0 -859.262810842738\n",
            "1 -1675.8493342618217\n",
            "2 -1069.7071672157972\n",
            "3 -1842.5100422050907\n",
            "4 -1763.8861050453704\n",
            "5 -1591.47847695624\n",
            "6 -743.1949531866867\n",
            "7 -1510.3625445096295\n",
            "8 -1477.451801086452\n",
            "9 -1074.3869481288978\n",
            "10 -1652.5485749991399\n",
            "11 -1824.9267533100017\n",
            "12 -1437.199775272934\n",
            "13 -1042.4618598043546\n",
            "14 -1517.4525642378646\n",
            "15 -1525.0239188651194\n",
            "16 -1726.9495491263176\n",
            "17 -1181.5594235281105\n",
            "18 -1037.2787479193853\n",
            "19 -992.2708222432041\n",
            "20 -1142.6840886090492\n",
            "21 -1787.0048468868818\n",
            "22 -888.1716269956055\n",
            "23 -1490.0798425397707\n",
            "24 -847.1468212981089\n",
            "25 -1587.3167705914866\n",
            "26 -1297.371041112738\n",
            "27 -1159.6777725796214\n",
            "28 -1262.7762030288516\n",
            "29 -1061.1944195978529\n",
            "30 -1146.4186298079874\n",
            "31 -871.7973815139381\n",
            "32 -1427.915130898322\n",
            "33 -963.5874073447183\n",
            "34 -1002.0820030953399\n",
            "35 -1179.8921993252045\n",
            "36 -1268.5852123658474\n",
            "37 -1164.5813602357136\n",
            "38 -1185.4194987098988\n",
            "39 -1344.8069306169932\n",
            "40 -1180.365489541117\n",
            "41 -1234.8773217388916\n",
            "42 -1338.8545694909276\n",
            "43 -1796.7205639576812\n",
            "44 -1331.0187509946766\n",
            "45 -1682.9356466430847\n",
            "46 -1144.114699510646\n",
            "47 -1044.595878520867\n",
            "48 -1284.9990032061012\n",
            "49 -1153.467393557147\n",
            "50 -1005.5504049846342\n",
            "51 -1148.13271620968\n",
            "52 -1802.6287473818809\n",
            "53 -1261.0301806282775\n",
            "54 -630.5191250977862\n",
            "55 -1105.7750222842778\n",
            "56 -747.4086818423689\n",
            "57 -595.7812281623786\n",
            "58 -861.7928684294843\n",
            "59 -1334.645608946331\n",
            "60 -1826.4643509755522\n",
            "61 -1126.773188131464\n",
            "62 -1192.3930145037434\n",
            "63 -1765.4747043341674\n",
            "64 -905.429855594528\n",
            "65 -931.9119053357884\n",
            "66 -1074.2169254452135\n",
            "67 -1204.7567316967761\n",
            "68 -898.889209175961\n",
            "69 -1387.1574577481636\n",
            "70 -1390.4255744394586\n",
            "71 -1092.3360993793774\n",
            "72 -1063.2810369137617\n",
            "73 -862.3162491582455\n",
            "74 -1610.676271160041\n",
            "model_changed..\n",
            "75 -1055.2576650264396\n",
            "76 -1657.0219953484607\n",
            "77 -1412.6623494397859\n",
            "78 -1046.8989022599142\n",
            "79 -1174.2201207376838\n",
            "80 -853.744445508413\n",
            "81 -1070.1304518648565\n",
            "82 -1260.5242108103291\n",
            "83 -1296.9618939616994\n",
            "84 -970.7060771111652\n",
            "85 -993.3665358053221\n",
            "86 -1296.5912975715378\n",
            "87 -969.5851441245619\n",
            "88 -1945.283237850845\n",
            "89 -1068.3995567253235\n",
            "90 -1638.599429127889\n",
            "91 -1045.1282015192878\n",
            "92 -1314.00730772714\n",
            "93 -1168.5497678744305\n",
            "94 -793.3057060665976\n",
            "95 -1446.7797298794453\n",
            "96 -993.8365131121852\n",
            "97 -1472.9440797628888\n",
            "98 -1320.8518774056588\n",
            "99 -1030.9184249611333\n",
            "100 -1447.9953357294057\n",
            "101 -1439.6263028862895\n",
            "102 -1157.214266542164\n",
            "103 -1452.971754054637\n",
            "104 -1428.9307859862404\n",
            "105 -1049.6982068044626\n",
            "106 -1057.6732576280067\n",
            "107 -1443.8949667358918\n",
            "108 -1521.039221226005\n",
            "109 -1869.108809836797\n",
            "110 -758.7103305756455\n",
            "111 -1836.4412845281672\n",
            "112 -1174.9835242050906\n",
            "113 -1920.043146937561\n",
            "114 -1281.2799292600314\n",
            "115 -757.2222271288053\n",
            "116 -1709.2850974063706\n",
            "117 -1080.2051072726072\n",
            "118 -898.798646273787\n",
            "119 -1600.4926713000848\n",
            "120 -748.1014400896535\n",
            "121 -1567.4306779897845\n",
            "122 -969.5775844603826\n",
            "123 -1367.3957558911577\n",
            "124 -1392.2150485932884\n",
            "125 -1752.575802582985\n",
            "126 -1506.1722401560241\n",
            "127 -1207.8815228726396\n",
            "128 -1234.3706120284155\n",
            "129 -1844.6247424096466\n",
            "130 -1729.2840689088596\n",
            "131 -973.2310126657655\n",
            "132 -969.7754115765042\n",
            "133 -1166.896149740371\n",
            "134 -1144.6879537901502\n",
            "135 -747.450966055705\n",
            "136 -1394.304724614339\n",
            "137 -1732.67269698888\n",
            "138 -1752.6850717444509\n",
            "139 -1876.6478237416495\n",
            "140 -1348.5979739929282\n",
            "141 -1285.8328898342218\n",
            "142 -1293.653344590017\n",
            "143 -1601.4134614941706\n",
            "144 -1403.2436928137377\n",
            "145 -1671.8258797125025\n",
            "146 -1471.405610023437\n",
            "147 -1654.0836683983455\n",
            "148 -1900.331701759984\n",
            "149 -1503.773448776637\n",
            "model_changed..\n",
            "150 -1457.715017583035\n",
            "151 -1602.6764112071749\n",
            "152 -1704.7191515964892\n",
            "153 -1473.958365561613\n",
            "154 -1160.6161264058926\n",
            "155 -755.7455679037693\n",
            "156 -863.6048067959398\n",
            "157 -968.7292752159237\n",
            "158 -1166.363617157559\n",
            "159 -1842.9902427870113\n",
            "160 -1861.070608431308\n",
            "161 -1579.3229251189377\n",
            "162 -973.02116457217\n",
            "163 -1058.2062582970223\n",
            "164 -1753.4944150563613\n",
            "165 -1068.4198303663002\n",
            "166 -1293.8371043413676\n",
            "167 -760.3619702968402\n",
            "168 -862.48392574717\n",
            "169 -1320.9262146282945\n",
            "170 -865.0812252776257\n",
            "171 -505.4062487088947\n",
            "172 -889.18548789404\n",
            "173 -1841.5737815964383\n",
            "174 -1468.1055612831738\n",
            "175 -1172.9075545147584\n",
            "176 -969.8753032406399\n",
            "177 -1224.7613515470593\n",
            "178 -1005.431529439996\n",
            "179 -964.134780207133\n",
            "180 -746.3348229378678\n",
            "181 -1403.683720690352\n",
            "182 -1325.4373100800585\n",
            "183 -975.1211198526544\n",
            "184 -1236.8227761490155\n",
            "185 -1310.5150992227616\n",
            "186 -1113.7621289948916\n",
            "187 -862.486726847646\n",
            "188 -1872.9804519440136\n",
            "189 -1451.636287192731\n",
            "190 -1886.3249520335266\n",
            "191 -1385.6811888831612\n",
            "192 -1740.7009443742302\n",
            "193 -1526.1454703329207\n",
            "194 -1156.9727210570384\n",
            "195 -979.0237234680382\n",
            "196 -1650.9573530662426\n",
            "197 -1385.8989668788342\n",
            "198 -1481.0670248363328\n",
            "199 -1326.0963843382544\n",
            "200 -641.1704140025888\n",
            "201 -819.1939700552507\n",
            "202 -651.9719987424556\n",
            "203 -1444.9998629248003\n",
            "204 -896.5724192602886\n",
            "205 -1156.366684569565\n",
            "206 -1410.2588433003493\n",
            "207 -1241.5703706698266\n",
            "208 -1728.3288962509846\n",
            "209 -1859.0986271167662\n",
            "210 -1169.1057970283737\n",
            "211 -744.4437961997197\n",
            "212 -1425.45015760417\n",
            "213 -1355.8606617533774\n",
            "214 -1096.0081809037488\n",
            "215 -1559.134641606072\n",
            "216 -1068.2411919642414\n",
            "217 -1863.9041858357061\n",
            "218 -1697.3882530469834\n",
            "219 -1697.374104471337\n",
            "220 -1423.272505828234\n",
            "221 -968.4144697841087\n",
            "222 -1385.333056757787\n",
            "223 -737.1489574453748\n",
            "224 -1483.1233912877308\n",
            "model_changed..\n",
            "225 -1617.5771871663667\n",
            "226 -762.6945982016398\n",
            "227 -1337.3737677293511\n",
            "228 -743.3146565399474\n",
            "229 -1134.4500643651663\n",
            "230 -1296.3584248881023\n",
            "231 -1361.8827225467896\n",
            "232 -1316.6673916364743\n",
            "233 -1167.8478842677284\n",
            "234 -1359.799933820525\n",
            "235 -751.4454090567042\n",
            "236 -961.1746030993419\n",
            "237 -1809.8275040463498\n",
            "238 -1695.0466340803227\n",
            "239 -966.7969541855178\n",
            "240 -861.7490653969545\n",
            "241 -953.0179221409322\n",
            "242 -1201.9809248046697\n",
            "243 -626.8972091597494\n",
            "244 -965.7116008123987\n",
            "245 -975.6185493673579\n",
            "246 -1329.901714642073\n",
            "247 -1791.6436873693233\n",
            "248 -961.5092197102607\n",
            "249 -1873.492014181169\n",
            "250 -870.4151301039956\n",
            "251 -1602.0496291769348\n",
            "252 -1070.2856663031064\n",
            "253 -1210.1477338763955\n",
            "254 -907.2769154815575\n",
            "255 -1858.098176060066\n",
            "256 -1584.3552868155878\n",
            "257 -1068.4356234191828\n",
            "258 -1863.299630909798\n",
            "259 -1735.5857311816392\n",
            "260 -1268.1105591826652\n",
            "261 -1554.380295054539\n",
            "262 -862.7828091497278\n",
            "263 -1160.473966757614\n",
            "264 -826.3775043084083\n",
            "265 -1663.173423494636\n",
            "266 -1345.5531212112485\n",
            "267 -1611.3811020111277\n",
            "268 -1661.5678074456068\n",
            "269 -1358.2292339814778\n",
            "270 -1375.942727492302\n",
            "271 -739.1122739018131\n",
            "272 -694.314508044372\n",
            "273 -1040.8132663651427\n",
            "274 -1701.815660913437\n",
            "275 -1844.7406368236852\n",
            "276 -1070.0686201465385\n",
            "277 -1565.2565936255203\n",
            "278 -1236.589810271561\n",
            "279 -1857.2898895274307\n",
            "280 -1125.656792299774\n",
            "281 -1664.0949925014\n",
            "282 -1231.807580589034\n",
            "283 -1483.0052997180824\n",
            "284 -1604.3857587893458\n",
            "285 -1766.8277370412197\n",
            "286 -1550.8969257582205\n",
            "287 -1069.8943296101947\n",
            "288 -1514.152357461794\n",
            "289 -1628.8445689282114\n",
            "290 -1551.886271281989\n",
            "291 -1434.406930820447\n",
            "292 -1762.6496002021793\n",
            "293 -1087.1702702789842\n",
            "294 -1453.2774642574934\n",
            "295 -1025.294523236588\n",
            "296 -1520.4732826226564\n",
            "297 -1301.9249833483184\n",
            "298 -1059.2408998108092\n",
            "299 -968.358083298415\n",
            "model_changed..\n",
            "300 -958.6082547737404\n",
            "301 -1500.9878150574068\n",
            "302 -939.6977676880744\n",
            "303 -1068.1275175886951\n",
            "304 -1036.4300895306342\n",
            "305 -864.2114010890074\n",
            "306 -761.0036157024472\n",
            "307 -969.8313595576169\n",
            "308 -1570.0002353263867\n",
            "309 -1145.7754184647629\n",
            "310 -1306.441873509365\n",
            "311 -766.0414494111621\n",
            "312 -1533.0785194887596\n",
            "313 -865.0231403602162\n",
            "314 -905.3505429675462\n",
            "315 -1287.5265356538935\n",
            "316 -1395.8171040265204\n",
            "317 -746.6380309585131\n",
            "318 -1734.6377119079332\n",
            "319 -997.0238884967307\n",
            "320 -898.6511759944788\n",
            "321 -1408.292583203807\n",
            "322 -1718.874221278552\n",
            "323 -862.7886718516537\n",
            "324 -1177.006750661681\n",
            "325 -970.4711137177856\n",
            "326 -763.0792010040152\n",
            "327 -1305.7792748063123\n",
            "328 -1603.4397463668074\n",
            "329 -1037.684161305105\n",
            "330 -1328.9531775522146\n",
            "331 -989.6701505833082\n",
            "332 -1822.8840489192737\n",
            "333 -741.4080346394015\n",
            "334 -906.2398423309255\n",
            "335 -1491.5674708776967\n",
            "336 -952.2581533156502\n",
            "337 -742.4500984076152\n",
            "338 -1446.3841748978064\n",
            "339 -1177.9373405086833\n",
            "340 -1474.7618442261025\n",
            "341 -1008.3069816644914\n",
            "342 -1017.6844671906453\n",
            "343 -1788.4951471709653\n",
            "344 -1250.345321046067\n",
            "345 -1182.6488570684696\n",
            "346 -1674.7095987074144\n",
            "347 -1273.3046043810389\n",
            "348 -1036.25717391026\n",
            "349 -1659.2661824705172\n",
            "350 -1070.2979949063279\n",
            "351 -1069.8869720424075\n",
            "352 -1114.5578209456628\n",
            "353 -1093.996056226349\n",
            "354 -1617.9941178538743\n",
            "355 -972.9158602554878\n",
            "356 -899.3572381050891\n",
            "357 -1284.3905201213995\n",
            "358 -1425.8900331556715\n",
            "359 -887.5461066479597\n",
            "360 -1481.015352129867\n",
            "361 -1104.8748596162807\n",
            "362 -862.1514083472149\n",
            "363 -1194.2256307003208\n",
            "364 -1306.4358327644657\n",
            "365 -901.7964004232199\n",
            "366 -861.5775434959756\n",
            "367 -1069.938281378079\n",
            "368 -970.1548210314861\n",
            "369 -1292.5677565929357\n",
            "370 -860.0548022027771\n",
            "371 -902.7488940169507\n",
            "372 -1008.1664729101183\n",
            "373 -1716.6768860106608\n",
            "374 -1381.0529867971638\n",
            "model_changed..\n",
            "375 -1319.3511059666766\n",
            "376 -1099.5482260885954\n",
            "377 -972.9326700267006\n",
            "378 -1788.897327462462\n",
            "379 -1186.9111030609283\n",
            "380 -1068.6195095244377\n",
            "381 -1289.1564580594934\n",
            "382 -859.3781246765378\n",
            "383 -1774.8746457146929\n",
            "384 -858.766581805436\n",
            "385 -922.3312469664332\n",
            "386 -871.098304964638\n",
            "387 -1241.6395322156661\n",
            "388 -960.0097670505378\n",
            "389 -998.4355782891365\n",
            "390 -964.5787199546556\n",
            "391 -1706.0989550025683\n",
            "392 -1301.9297804518358\n",
            "393 -1165.654606432125\n",
            "394 -1358.065504142207\n",
            "395 -1604.6286390046341\n",
            "396 -1140.7774583026128\n",
            "397 -1482.78664446524\n",
            "398 -1428.5526220971378\n",
            "399 -1703.7998704428492\n",
            "400 -1247.6402136706756\n",
            "401 -1275.2400085265056\n",
            "402 -1082.1767433608784\n",
            "403 -1369.5145135561431\n",
            "404 -1502.850269818116\n",
            "405 -1400.8248128669075\n",
            "406 -1600.9658202275198\n",
            "407 -1351.9637987938445\n",
            "408 -1328.6807812144882\n",
            "409 -1192.8059918934393\n",
            "410 -1277.1347403882378\n",
            "411 -1299.9585096650346\n",
            "412 -1451.2637776473105\n",
            "413 -1318.8630201773724\n",
            "414 -1790.897312699603\n",
            "415 -1575.5982070092296\n",
            "416 -1274.9243487961608\n",
            "417 -1486.1348644020384\n",
            "418 -1426.9067551794592\n",
            "419 -1269.6819300262302\n",
            "420 -1196.361800846756\n",
            "421 -1471.8195647530656\n",
            "422 -1212.9362772448037\n",
            "423 -1436.5753700775094\n",
            "424 -1068.7171702612993\n",
            "425 -1281.6348894350765\n",
            "426 -830.4207594760146\n",
            "427 -1269.9609621333764\n",
            "428 -826.6576787446803\n",
            "429 -1521.1401617877953\n",
            "430 -1165.9457282571127\n",
            "431 -860.5649786448442\n",
            "432 -1145.4562002720406\n",
            "433 -1567.7453576473758\n",
            "434 -1703.2789137413401\n",
            "435 -974.2333066447519\n",
            "436 -1807.7908914121447\n",
            "437 -1066.4277648743503\n",
            "438 -1176.3003410033286\n",
            "439 -923.8402099920837\n",
            "440 -1462.5745701988546\n",
            "441 -1289.800062677379\n",
            "442 -1722.9381783652495\n",
            "443 -1271.6432735059961\n",
            "444 -938.4554012532467\n",
            "445 -1148.8676677313608\n",
            "446 -1172.4873561320123\n",
            "447 -1369.347598692903\n",
            "448 -975.1585758643279\n",
            "449 -1066.6184090371278\n",
            "model_changed..\n",
            "450 -969.4741991270425\n",
            "451 -1691.9589455835082\n",
            "452 -899.6408656200508\n",
            "453 -1264.913377414632\n",
            "454 -747.4329374654139\n",
            "455 -1785.8352097937322\n",
            "456 -1276.1287704290914\n",
            "457 -1202.7063974562116\n",
            "458 -1915.3693276773472\n",
            "459 -1614.0731425107804\n",
            "460 -1155.279649653843\n",
            "461 -966.7141228973718\n",
            "462 -1074.0324489707702\n",
            "463 -934.866928128488\n",
            "464 -897.2354068614613\n",
            "465 -1273.382982004608\n",
            "466 -1728.4100627060614\n",
            "467 -1308.1129818828588\n",
            "468 -1843.0232424708097\n",
            "469 -856.2989579831676\n",
            "470 -1066.278952897321\n",
            "471 -1760.5119400284364\n",
            "472 -1318.6006704346178\n",
            "473 -862.7527489086833\n",
            "474 -1073.547666880483\n",
            "475 -965.462237710422\n",
            "476 -1379.72804238951\n",
            "477 -1907.660332678238\n",
            "478 -1287.3809192671802\n",
            "479 -910.3252808339889\n",
            "480 -1088.591788410098\n",
            "481 -1342.305263725168\n",
            "482 -1221.8498396386149\n",
            "483 -1054.7337289973354\n",
            "484 -1093.6736190939037\n",
            "485 -1185.6767101131684\n",
            "486 -1458.6120905805328\n",
            "487 -1763.8537694620609\n",
            "488 -1156.55909105926\n",
            "489 -743.8802974450278\n",
            "490 -1201.3093760746585\n",
            "491 -998.8040208144636\n",
            "492 -1175.4680768187159\n",
            "493 -969.3948265967999\n",
            "494 -1568.3119092354864\n",
            "495 -1180.6799206761839\n",
            "496 -1825.1875744193876\n",
            "497 -745.8024902164296\n",
            "498 -1104.5079750511843\n",
            "499 -1745.4021686019307\n",
            "500 -1049.5952670681443\n",
            "501 -1353.968930721077\n",
            "502 -704.882880203021\n",
            "503 -1602.429428114515\n",
            "504 -1300.6646503344\n",
            "505 -1453.3891470013305\n",
            "506 -825.1717937169286\n",
            "507 -1667.7766646021166\n",
            "508 -1299.9719498498478\n",
            "509 -1693.8145785403035\n",
            "510 -965.9422625560454\n",
            "511 -1070.1844886582123\n",
            "512 -1381.893260217926\n",
            "513 -1210.2404104142984\n",
            "514 -1697.4935025690502\n",
            "515 -1252.8807920552342\n",
            "516 -1070.9169773178655\n",
            "517 -770.6297942552437\n",
            "518 -1356.839528874189\n",
            "519 -1487.7076885777296\n",
            "520 -1171.4886809841034\n",
            "521 -1374.8619095634797\n",
            "522 -1675.4283956287204\n",
            "523 -895.0755303855447\n",
            "524 -1527.7749284212514\n",
            "model_changed..\n",
            "525 -1432.451862432634\n",
            "526 -1798.3144070343942\n",
            "527 -1248.6088660311177\n",
            "528 -1694.651015086356\n",
            "529 -860.702898032321\n",
            "530 -969.1745507821133\n",
            "531 -1121.8301515152673\n",
            "532 -793.067335970405\n",
            "533 -970.3202877331858\n",
            "534 -937.6948996720956\n",
            "535 -1954.2599713737059\n",
            "536 -1396.9062120721653\n",
            "537 -1084.8483971016265\n",
            "538 -891.5858798149784\n",
            "539 -941.7853003770108\n",
            "540 -1174.651593194014\n",
            "541 -1280.484785295571\n",
            "542 -1600.6854037305159\n",
            "543 -1071.1403330685541\n",
            "544 -1262.3272328666687\n",
            "545 -1313.774562813343\n",
            "546 -898.4061601781344\n",
            "547 -900.5826682989833\n",
            "548 -971.8114869029816\n",
            "549 -1604.5498045108577\n",
            "550 -1000.6799551796101\n",
            "551 -904.162300101431\n",
            "552 -1071.3944981456627\n",
            "553 -1156.1128160372073\n",
            "554 -1652.3750184746448\n",
            "555 -1591.3255130501866\n",
            "556 -799.9952955717386\n",
            "557 -1918.3661256820046\n",
            "558 -1746.8437241795057\n",
            "559 -887.724356605743\n",
            "560 -1696.2145305228457\n",
            "561 -1034.7344311166737\n",
            "562 -1255.0330454434106\n",
            "563 -1020.2419745134323\n",
            "564 -1299.0365308422108\n",
            "565 -1369.052775463328\n",
            "566 -849.8790082983603\n",
            "567 -1072.238375825085\n",
            "568 -1173.1644334414054\n",
            "569 -1313.8010892954649\n",
            "570 -1452.4571958382555\n",
            "571 -1657.2230945695574\n",
            "572 -1276.216900178321\n",
            "573 -1309.2312396239197\n",
            "574 -1113.7808784249814\n",
            "575 -969.9267459894425\n",
            "576 -1419.0420446871879\n",
            "577 -1054.0679177232764\n",
            "578 -1173.5319320349495\n",
            "579 -1351.9768097949607\n",
            "580 -1618.7501930638412\n",
            "581 -1089.9118931199446\n",
            "582 -1068.8301584978633\n",
            "583 -1069.35904949936\n",
            "584 -1014.8870241600627\n",
            "585 -1041.3443698109095\n",
            "586 -1028.6278011927848\n",
            "587 -1040.1739496365578\n",
            "588 -1791.4074843196468\n",
            "589 -1931.384238647607\n",
            "590 -1023.898311930227\n",
            "591 -891.6255212111378\n",
            "592 -1786.3518194804653\n",
            "593 -968.9095118002805\n",
            "594 -1042.779840798208\n",
            "595 -1036.827763684045\n",
            "596 -1058.994006955639\n",
            "597 -1072.1006397192266\n",
            "598 -1165.760306286299\n",
            "599 -967.9368094166463\n",
            "model_changed..\n",
            "600 -1475.0642259959968\n",
            "601 -1297.9823096691828\n",
            "602 -1465.9870106170936\n",
            "603 -1030.615246642863\n",
            "604 -1422.7624547860876\n",
            "605 -770.8019004039493\n",
            "606 -1718.6860827281844\n",
            "607 -1259.4787973914547\n",
            "608 -1526.3627050926941\n",
            "609 -1426.7117891519188\n",
            "610 -854.0490040722169\n",
            "611 -1165.4505415553008\n",
            "612 -1098.159708503467\n",
            "613 -1509.034559673456\n",
            "614 -1879.0733659450214\n",
            "615 -1597.1591791327646\n",
            "616 -1610.2085193657017\n",
            "617 -1190.4439341230127\n",
            "618 -1144.3179309970578\n",
            "619 -1197.6273343957957\n",
            "620 -1077.1203334920033\n",
            "621 -960.1161223477961\n",
            "622 -1206.6085479787112\n",
            "623 -957.8686474270664\n",
            "624 -1152.2706995016235\n",
            "625 -1725.9072087728\n",
            "626 -1374.3206087304459\n",
            "627 -1531.927908937254\n",
            "628 -1199.1058458488935\n",
            "629 -1346.1937706671092\n",
            "630 -1748.1572467856024\n",
            "631 -1542.1818490868632\n",
            "632 -1284.3720039449104\n",
            "633 -1200.3430186919165\n",
            "634 -1320.0905810424165\n",
            "635 -1832.4811400340914\n",
            "636 -1164.4225884267948\n",
            "637 -1242.467491359005\n",
            "638 -1021.0810778701245\n",
            "639 -1303.3224092886203\n",
            "640 -967.7160399480448\n",
            "641 -1669.3706153537005\n",
            "642 -1166.335419460635\n",
            "643 -1017.6625795674927\n",
            "644 -1200.9517102603\n",
            "645 -1186.8823267566015\n",
            "646 -1065.4420449463316\n",
            "647 -1139.3961885662052\n",
            "648 -1250.544543097454\n",
            "649 -1116.819708230853\n",
            "650 -1292.3331668995397\n",
            "651 -1665.4193778453543\n",
            "652 -849.7343836409246\n",
            "653 -1064.7282207746973\n",
            "654 -1215.467168557035\n",
            "655 -1455.1864583816373\n",
            "656 -954.2108102989827\n",
            "657 -1252.591150237603\n",
            "658 -1744.3338562900312\n",
            "659 -1670.5968127307938\n",
            "660 -1528.6928647236891\n",
            "661 -1206.2205419437112\n",
            "662 -1415.170817082236\n",
            "663 -1087.666079020149\n",
            "664 -1485.9756004816604\n",
            "665 -1571.5894965810562\n",
            "666 -1340.8903209139828\n",
            "667 -1168.6247028840546\n",
            "668 -1091.2746473355369\n",
            "669 -1140.4523674278905\n",
            "670 -1158.7373308342712\n",
            "671 -1353.1188749127452\n",
            "672 -1658.8493973341153\n",
            "673 -1591.2732180644941\n",
            "674 -1123.6935412564974\n",
            "model_changed..\n",
            "675 -1467.3874592429468\n",
            "676 -1612.2727843558348\n",
            "677 -1838.5283037131148\n",
            "678 -1177.1179978277687\n",
            "679 -1069.2082337646636\n",
            "680 -1269.103671422955\n",
            "681 -1113.53962646946\n",
            "682 -1843.6932771527609\n",
            "683 -882.7983773568968\n",
            "684 -1050.0726654665073\n",
            "685 -1509.7492071520855\n",
            "686 -1348.6566874618306\n",
            "687 -1427.682794662929\n",
            "688 -1712.7423766006568\n",
            "689 -1054.0360482774552\n",
            "690 -975.8124137945156\n",
            "691 -862.4158443543025\n",
            "692 -1581.9557639949642\n",
            "693 -935.1215028759192\n",
            "694 -959.4053477748009\n",
            "695 -1558.3478198672751\n",
            "696 -1309.0968871200523\n",
            "697 -929.1345403292345\n",
            "698 -1168.9556481801403\n",
            "699 -1400.1888342046184\n",
            "700 -1734.8597853949416\n",
            "701 -1390.6296382012517\n",
            "702 -1818.9818607261454\n",
            "703 -1508.8394796773682\n",
            "704 -1437.6459497165704\n",
            "705 -1171.5693461890487\n",
            "706 -1368.3937567332541\n",
            "707 -907.5226843939022\n",
            "708 -1026.6277599302182\n",
            "709 -1661.258489758819\n",
            "710 -1761.394561419609\n",
            "711 -1846.399438375668\n",
            "712 -1513.4082026900198\n",
            "713 -1273.2075107929843\n",
            "714 -1031.88940246641\n",
            "715 -1307.5087465172378\n",
            "716 -1188.5052247758194\n",
            "717 -1042.4174928061964\n",
            "718 -1526.6229595670413\n",
            "719 -1066.914786575139\n",
            "720 -1054.0814737754304\n",
            "721 -1319.8970725787638\n",
            "722 -1458.2648240313204\n",
            "723 -965.4784777135873\n",
            "724 -1746.8134541640038\n",
            "725 -1098.1611371860317\n",
            "726 -1058.5130662182212\n",
            "727 -1878.7278147349223\n",
            "728 -1256.7473991007082\n",
            "729 -1292.4789336035417\n",
            "730 -1161.2199763510307\n",
            "731 -861.7988048710015\n",
            "732 -1572.5592548306875\n",
            "733 -1329.57643361895\n",
            "734 -1324.0223893484886\n",
            "735 -1799.4151379374396\n",
            "736 -1069.4987160177382\n",
            "737 -1800.9382201486824\n",
            "738 -1359.0635783945\n",
            "739 -1551.8850938726534\n",
            "740 -1038.4606391025673\n",
            "741 -964.7558477363445\n",
            "742 -882.3993491068128\n",
            "743 -1316.2784870941061\n",
            "744 -1258.8201028682524\n",
            "745 -1747.2199359573465\n",
            "746 -1261.4286650823894\n",
            "747 -1797.368474575197\n",
            "748 -1715.979382083459\n",
            "749 -1451.3044534578519\n",
            "model_changed..\n",
            "750 -1357.3730355138562\n",
            "751 -1364.5144822507434\n",
            "752 -1071.746900353409\n",
            "753 -1777.1312750936356\n",
            "754 -967.458358922848\n",
            "755 -1339.981062301047\n",
            "756 -1504.3477286247703\n",
            "757 -1252.7195762523404\n",
            "758 -1280.0626401769423\n",
            "759 -958.809237026633\n",
            "760 -984.3561036887638\n",
            "761 -1266.0038901269518\n",
            "762 -1318.357713047472\n",
            "763 -1288.6278513837153\n",
            "764 -1147.2884567915569\n",
            "765 -1203.5478152001156\n",
            "766 -857.4005034827647\n",
            "767 -749.0344565006533\n",
            "768 -1424.2515256658342\n",
            "769 -1656.3819694020117\n",
            "770 -1208.8013713854057\n",
            "771 -856.1246823068037\n",
            "772 -1019.1328071358105\n",
            "773 -1284.8107580513501\n",
            "774 -1174.986272955282\n",
            "775 -930.1508362672059\n",
            "776 -1484.5620100433282\n",
            "777 -1299.204548959139\n",
            "778 -1614.1329614950873\n",
            "779 -1037.3045054272557\n",
            "780 -1346.9329187644632\n",
            "781 -1160.8667898179544\n",
            "782 -1712.6771539629308\n",
            "783 -1314.157387447573\n",
            "784 -1193.8421919164807\n",
            "785 -1644.4426232993137\n",
            "786 -1218.27345033797\n",
            "787 -1583.1413406265476\n",
            "788 -1578.3664307522697\n",
            "789 -1190.298161203542\n",
            "790 -948.0261484141043\n",
            "791 -1229.75467412967\n",
            "792 -958.6490511015411\n",
            "793 -1127.900458033975\n",
            "794 -965.9057932115411\n",
            "795 -1515.3367655171387\n",
            "796 -1008.5297824854216\n",
            "797 -1439.7738988067626\n",
            "798 -1170.7136369579025\n",
            "799 -1390.005791758278\n",
            "800 -1191.370559573621\n",
            "801 -1177.187607174542\n",
            "802 -839.6501101724405\n",
            "803 -1805.156364684415\n",
            "804 -1170.7007610092905\n",
            "805 -1215.7812157063763\n",
            "806 -1197.9302765321254\n",
            "807 -1223.6460616544173\n",
            "808 -1241.44091670088\n",
            "809 -1264.5784053849447\n",
            "810 -949.1503792386972\n",
            "811 -1297.4844827257905\n",
            "812 -1289.463236552213\n",
            "813 -1317.3097919085553\n",
            "814 -1730.9573266113744\n",
            "815 -1540.1715105602618\n",
            "816 -1422.249074351366\n",
            "817 -1446.4318023271876\n",
            "818 -1063.2361138544518\n",
            "819 -1398.1695925349532\n",
            "820 -1106.0127833822162\n",
            "821 -1181.6724941167467\n",
            "822 -1180.1016746386479\n",
            "823 -1281.305933370929\n",
            "824 -944.2916390549861\n",
            "model_changed..\n",
            "825 -1791.4882481052596\n",
            "826 -1153.1126912223667\n",
            "827 -1177.0002023818436\n",
            "828 -1178.0974310324216\n",
            "829 -1501.0656133187508\n",
            "830 -1339.7168692679265\n",
            "831 -1118.452577142626\n",
            "832 -1351.0764076331943\n",
            "833 -1252.4393128812262\n",
            "834 -1327.4523774380261\n",
            "835 -1456.2916574605388\n",
            "836 -1211.7909684956653\n",
            "837 -1337.64779929203\n",
            "838 -1483.3910977348899\n",
            "839 -1160.7591724049696\n",
            "840 -1101.0980114070758\n",
            "841 -1432.6705323774506\n",
            "842 -1624.62072893897\n",
            "843 -1051.2297681565258\n",
            "844 -1196.9686860112724\n",
            "845 -1023.8373687428988\n",
            "846 -992.6172534193649\n",
            "847 -1572.0676886715846\n",
            "848 -1171.6434993791288\n",
            "849 -1610.5086624598052\n",
            "850 -1592.0401698140477\n",
            "851 -1349.6446046099582\n",
            "852 -1034.5594897833203\n",
            "853 -1212.0076794478293\n",
            "854 -1169.941308016123\n",
            "855 -1203.7952003924627\n",
            "856 -1511.4716491858474\n",
            "857 -1069.7854581896845\n",
            "858 -1303.3525215587485\n",
            "859 -1578.3934585466754\n",
            "860 -1196.299393568218\n",
            "861 -1011.7451030492714\n",
            "862 -1289.4700656554212\n",
            "863 -1390.6304844327408\n",
            "864 -1140.0216793276015\n",
            "865 -1181.7122583616201\n",
            "866 -1329.5717840455488\n",
            "867 -1179.3202230632855\n",
            "868 -1316.6657273185144\n",
            "869 -1227.027371689507\n",
            "870 -1665.2087484277365\n",
            "871 -1429.6133692885599\n",
            "872 -1604.4863069398814\n",
            "873 -1595.2501178689613\n",
            "874 -1190.3247454878403\n",
            "875 -1541.5976179762788\n",
            "876 -1307.3608629337934\n",
            "877 -1627.3067994895023\n",
            "878 -1362.3049049771082\n",
            "879 -1843.2437107074898\n",
            "880 -1185.2439706203022\n",
            "881 -1379.1415472177255\n",
            "882 -1640.9118923939793\n",
            "883 -1187.0416880505186\n",
            "884 -1161.6343758422656\n",
            "885 -1270.4238866491771\n",
            "886 -1757.3333128018862\n",
            "887 -1114.8252267065645\n",
            "888 -1547.265301333049\n",
            "889 -1578.5372814877603\n",
            "890 -1155.6897365631478\n",
            "891 -1525.4940440214025\n",
            "892 -1030.6879149952824\n",
            "893 -1086.311216233168\n",
            "894 -1184.0534235049786\n",
            "895 -1065.8189504127033\n",
            "896 -1529.0963535386\n",
            "897 -1076.2219545881578\n",
            "898 -1463.0497008509328\n",
            "899 -1828.3794337327504\n",
            "model_changed..\n",
            "900 -1833.1169205233784\n",
            "901 -1168.5354612490055\n",
            "902 -853.6676590900777\n",
            "903 -1075.771181242424\n",
            "904 -740.0267748674812\n",
            "905 -1381.7959027380873\n",
            "906 -957.2660183689451\n",
            "907 -835.6351206754316\n",
            "908 -1468.3769353011555\n",
            "909 -1171.9650787663038\n",
            "910 -978.1445530785356\n",
            "911 -1269.0882813296814\n",
            "912 -730.5253615697972\n",
            "913 -1033.274397971026\n",
            "914 -1821.4363618172526\n",
            "915 -1108.636139320989\n",
            "916 -1595.414408219883\n",
            "917 -963.1677200203651\n",
            "918 -798.5628790861525\n",
            "919 -1594.087170666229\n",
            "920 -1757.6134644915676\n",
            "921 -1351.4518983656244\n",
            "922 -1708.6349603608758\n",
            "923 -1726.9197726478242\n",
            "924 -1784.4073879697041\n",
            "925 -916.7247588214591\n",
            "926 -1176.6056135443944\n",
            "927 -1233.091757305691\n",
            "928 -1172.4573987604213\n",
            "929 -1163.1559843008502\n",
            "930 -1778.1346018044808\n",
            "931 -972.2029440942536\n",
            "932 -1747.6992438775114\n",
            "933 -1774.8959852873118\n",
            "934 -1720.1672505155307\n",
            "935 -968.7706962095673\n",
            "936 -1068.5628363158119\n",
            "937 -1177.813575415841\n",
            "938 -1858.1300417477503\n",
            "939 -1328.4878556086094\n",
            "940 -629.8601829675305\n",
            "941 -1384.4416806091463\n",
            "942 -857.2160576446414\n",
            "943 -1149.7682977237876\n",
            "944 -1188.2927142536\n",
            "945 -1058.73862799435\n",
            "946 -747.1619541583027\n",
            "947 -1279.8816942191208\n",
            "948 -1497.6506629172393\n",
            "949 -966.7829101890327\n",
            "950 -1697.8740584853074\n",
            "951 -1041.0354409407182\n",
            "952 -1687.86163744562\n",
            "953 -1447.2025059381047\n",
            "954 -1651.7355125458953\n",
            "955 -1592.23403202228\n",
            "956 -1743.853407699158\n",
            "957 -1008.5178286236273\n",
            "958 -1071.1539628754267\n",
            "959 -1336.073021080027\n",
            "960 -948.9116151024393\n",
            "961 -1118.2227529480454\n",
            "962 -961.4094928972675\n",
            "963 -861.2037441948396\n",
            "964 -900.1990495798483\n",
            "965 -1406.238093698616\n",
            "966 -1040.0448201416723\n",
            "967 -1229.6201273086451\n",
            "968 -1690.4099033433417\n",
            "969 -987.3908996263384\n",
            "970 -1455.8937285577736\n",
            "971 -1490.4246775985546\n",
            "972 -1151.7784043797606\n",
            "973 -939.7871257884136\n",
            "974 -1168.1743532800335\n",
            "model_changed..\n",
            "975 -1083.3898775773566\n",
            "976 -1458.2191991079628\n",
            "977 -1008.2723424161229\n",
            "978 -1624.6947761351869\n",
            "979 -1449.3232918163824\n",
            "980 -1064.3476708606538\n",
            "981 -1698.2355880133593\n",
            "982 -1626.4997113152908\n",
            "983 -756.024763864632\n",
            "984 -1011.0569657126746\n",
            "985 -1161.4623005991853\n",
            "986 -1036.6625345219973\n",
            "987 -969.6099092437496\n",
            "988 -1299.1049597099848\n",
            "989 -1540.5764921800535\n",
            "990 -1615.759562673358\n",
            "991 -1483.9590550799085\n",
            "992 -1065.4099924458747\n",
            "993 -1123.0899929569844\n",
            "994 -992.3017424998501\n",
            "995 -1168.7508294183976\n",
            "996 -982.0926600061126\n",
            "997 -904.2634651749821\n",
            "998 -1014.7641066820742\n",
            "999 -1184.5120358163672\n",
            "1000 -758.9476448952365\n",
            "1001 -1189.1499895090337\n",
            "1002 -1164.3549298471116\n",
            "1003 -1415.5703866711317\n",
            "1004 -965.35316854472\n",
            "1005 -1211.2151835698742\n",
            "1006 -1171.3266154926364\n",
            "1007 -1081.7593954382812\n",
            "1008 -1514.615517108659\n",
            "1009 -1177.5655105420792\n",
            "1010 -1395.4532189398788\n",
            "1011 -951.5604285242264\n",
            "1012 -1066.3467672387012\n",
            "1013 -1671.5842902204542\n",
            "1014 -772.411419666624\n",
            "1015 -1665.2498793591574\n",
            "1016 -1114.4772659778557\n",
            "1017 -1088.2039634942275\n",
            "1018 -1064.681345912532\n",
            "1019 -975.103124882295\n",
            "1020 -1042.043041621271\n",
            "1021 -969.9305768339386\n",
            "1022 -1536.0682510990114\n",
            "1023 -1219.713618813788\n",
            "1024 -1318.6957681764484\n",
            "1025 -1069.6613946405805\n",
            "1026 -1739.0320984830673\n",
            "1027 -1181.4380644275057\n",
            "1028 -1059.671718466389\n",
            "1029 -1558.3916506676405\n",
            "1030 -1690.8082276167986\n",
            "1031 -1235.6313046109008\n",
            "1032 -1240.8714320598287\n",
            "1033 -1356.9463128721186\n",
            "1034 -1071.005825758035\n",
            "1035 -1504.8057550572883\n",
            "1036 -1419.0053565740036\n",
            "1037 -1673.6538922882207\n",
            "1038 -892.7800471293166\n",
            "1039 -1066.294036081951\n",
            "1040 -861.6374347135586\n",
            "1041 -1264.56526461936\n",
            "1042 -1641.5847265833072\n",
            "1043 -1032.326863549916\n",
            "1044 -965.3601729419192\n",
            "1045 -1044.360624730182\n",
            "1046 -1594.7724556309786\n",
            "1047 -1599.8387789283981\n",
            "1048 -1744.4287188288095\n",
            "1049 -1069.3705426190181\n",
            "model_changed..\n",
            "1050 -1046.6061463864905\n",
            "1051 -1162.0389117627087\n",
            "1052 -902.8532607325245\n",
            "1053 -1344.7159356064983\n",
            "1054 -1053.5604249803446\n",
            "1055 -1579.7581586824945\n",
            "1056 -1150.383835720401\n",
            "1057 -1070.1654523011161\n",
            "1058 -964.7182216143585\n",
            "1059 -1336.497878842865\n",
            "1060 -1345.6730837380285\n",
            "1061 -1162.9124666752593\n",
            "1062 -1371.903368288253\n",
            "1063 -1769.0543305585288\n",
            "1064 -1457.3750499403866\n",
            "1065 -1066.3854211147802\n",
            "1066 -1148.0337378036402\n",
            "1067 -894.506107995061\n",
            "1068 -1770.8490148696367\n",
            "1069 -1514.6977347794311\n",
            "1070 -1312.6473828550318\n",
            "1071 -1073.1367408570986\n",
            "1072 -1631.2986474632817\n",
            "1073 -1573.0527502195782\n",
            "1074 -1524.360804982492\n",
            "1075 -900.64662575612\n",
            "1076 -1030.3412581976572\n",
            "1077 -1287.4369297954577\n",
            "1078 -1304.0893511064205\n",
            "1079 -1453.1319548368615\n",
            "1080 -1594.8062361975453\n",
            "1081 -1498.9436402724125\n",
            "1082 -1299.5226289261109\n",
            "1083 -1200.0511313004354\n",
            "1084 -1062.9326645653875\n",
            "1085 -1290.4196602262994\n",
            "1086 -965.501560811399\n",
            "1087 -1497.9947920681066\n",
            "1088 -1579.1554810094883\n",
            "1089 -1278.791007661544\n",
            "1090 -1297.6722291330352\n",
            "1091 -847.9809779343907\n",
            "1092 -1293.5343575069085\n",
            "1093 -1186.459473889291\n",
            "1094 -1444.5880662964444\n",
            "1095 -1096.4797986686244\n",
            "1096 -1138.9158328120784\n",
            "1097 -720.7960595641348\n",
            "1098 -1356.1293946064982\n",
            "1099 -1355.5073832183039\n",
            "1100 -1270.2075083358147\n",
            "1101 -848.9076607112975\n",
            "1102 -1500.0599276676523\n",
            "1103 -1509.3950971266725\n",
            "1104 -1331.8786146889931\n",
            "1105 -1751.317237147537\n",
            "1106 -1336.8307095187456\n",
            "1107 -1373.8443945750716\n",
            "1108 -1397.0210354327864\n",
            "1109 -1250.990624157523\n",
            "1110 -1052.509683897187\n",
            "1111 -963.3322318444039\n",
            "1112 -1826.0944641897565\n",
            "1113 -945.0851882730532\n",
            "1114 -1307.9018174756823\n",
            "1115 -1323.8741162973283\n",
            "1116 -1255.1982090850233\n",
            "1117 -1563.1315591271436\n",
            "1118 -1423.4558340271565\n",
            "1119 -1187.814212721044\n",
            "1120 -1602.4888501471646\n",
            "1121 -957.5673716791312\n",
            "1122 -1768.2109039814547\n",
            "1123 -1310.7169353405532\n",
            "1124 -1279.6234480603498\n",
            "model_changed..\n",
            "1125 -1299.8696155828436\n",
            "1126 -1481.3561606526462\n",
            "1127 -1449.827968039248\n",
            "1128 -1296.9550593012698\n",
            "1129 -1334.1526574168454\n",
            "1130 -1785.407389595986\n",
            "1131 -1066.2984769504394\n",
            "1132 -854.6155043052555\n",
            "1133 -1158.861341472808\n",
            "1134 -1405.0329053205855\n",
            "1135 -1205.1514677225114\n",
            "1136 -1766.9806368162356\n",
            "1137 -1623.4861802141952\n",
            "1138 -1352.879626003944\n",
            "1139 -1746.1083312113365\n",
            "1140 -1553.028105658693\n",
            "1141 -1279.3306064762514\n",
            "1142 -1748.5771469343729\n",
            "1143 -1047.8866044568204\n",
            "1144 -1531.562324007264\n",
            "1145 -1315.0488799045556\n",
            "1146 -1173.8711540073768\n",
            "1147 -1215.2192728860455\n",
            "1148 -1202.0482224450975\n",
            "1149 -1045.5757793888424\n",
            "1150 -1065.9757300308686\n",
            "1151 -1022.0120610849041\n",
            "1152 -1046.1478561667598\n",
            "1153 -1862.9190909857323\n",
            "1154 -1157.6081945282772\n",
            "1155 -813.9214663865145\n",
            "1156 -1356.4796539703198\n",
            "1157 -902.343649083653\n",
            "1158 -1738.8106443642364\n",
            "1159 -1297.7836234570277\n",
            "1160 -1062.4150753473239\n",
            "1161 -1690.277794950333\n",
            "1162 -1167.9830119634312\n",
            "1163 -1289.5932820647067\n",
            "1164 -1510.1024323579834\n",
            "1165 -1203.0077824794582\n",
            "1166 -1496.2744989462324\n",
            "1167 -735.1120078543373\n",
            "1168 -1413.7169415885164\n",
            "1169 -1653.454625194451\n",
            "1170 -958.6656688721066\n",
            "1171 -834.0319105620501\n",
            "1172 -1164.1069661927836\n",
            "1173 -952.4673322491547\n",
            "1174 -1431.0148053227088\n",
            "1175 -1080.0962592943386\n",
            "1176 -1561.4234797245251\n",
            "1177 -1631.4055550004016\n",
            "1178 -1593.2491249121128\n",
            "1179 -1647.6670160408476\n",
            "1180 -1325.884892009219\n",
            "1181 -1184.0925271073424\n",
            "1182 -1646.5740956215966\n",
            "1183 -1623.2028754429034\n",
            "1184 -803.6095090821611\n",
            "1185 -1615.0774901401069\n",
            "1186 -1650.387387326955\n",
            "1187 -1492.6036017122435\n",
            "1188 -1048.409753736054\n",
            "1189 -1629.1166024682004\n",
            "1190 -1613.6686146370932\n",
            "1191 -1654.8955804460031\n",
            "1192 -1516.2007901510053\n",
            "1193 -1542.7946708935863\n",
            "1194 -1653.9125981809502\n",
            "1195 -1526.8267405773588\n",
            "1196 -1617.7332333064708\n",
            "1197 -1090.5115170284896\n",
            "1198 -1496.8039428381473\n",
            "1199 -1325.239816261631\n"
          ]
        },
        {
          "output_type": "error",
          "ename": "AttributeError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-8-d312c29dfc5d>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     45\u001b[0m \u001b[0mfig1loss\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0max1loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msubplots\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     46\u001b[0m \u001b[0mx_loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0marange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcritic_loss_list\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---> 47\u001b[0;31m \u001b[0mcritic_loss_list\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcpu\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     48\u001b[0m \u001b[0max1loss\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx_loss\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mcritic_loss_list\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     49\u001b[0m \u001b[0max1loss\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mset_xlabel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'steps'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mAttributeError\u001b[0m: 'list' object has no attribute 'cpu'"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "fig1,ax1 = plt.subplots()\n",
        "x = np.arange(len(rew_list))\n",
        "ax1.plot(x,rew_list)\n",
        "ax1.set_xlabel('episodes')\n",
        "ax1.set_ylabel('rewards')\n",
        "fig1loss,ax1loss = plt.subplots()\n",
        "x_loss = np.arange(len(critic_loss_list))\n",
        "ax1loss.plot(x_loss,critic_loss_list)\n",
        "ax1loss.set_xlabel('steps')\n",
        "ax1loss.set_ylabel('loss')\n",
        "fig_al, ax_al = plt.subplots()\n",
        "x_aloss = np.arange(len(actor_loss_list))\n",
        "ax_al.plot(x_aloss,actor_loss_list)\n",
        "ax_al.set_xlabel('steps')\n",
        "ax_al.set_ylabel('loss')\n",
        "\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "3cn4oYkDfmek",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 929
        },
        "outputId": "e1d8fe45-d20a-435e-d0b4-2c7aee01c9e6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "TypeError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-9-4edb44fc9c26>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      6\u001b[0m \u001b[0mfig1loss\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0max1loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msubplots\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      7\u001b[0m \u001b[0mx_loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0marange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcritic_loss_list\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----> 8\u001b[0;31m \u001b[0max1loss\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx_loss\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mcritic_loss_list\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      9\u001b[0m \u001b[0max1loss\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mset_xlabel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'steps'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     10\u001b[0m \u001b[0max1loss\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mset_ylabel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'loss'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/matplotlib/axes/_axes.py\u001b[0m in \u001b[0;36mplot\u001b[0;34m(self, scalex, scaley, data, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1645\u001b[0m         \"\"\"\n\u001b[1;32m   1646\u001b[0m         \u001b[0mkwargs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcbook\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnormalize_kwargs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmlines\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mLine2D\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1647\u001b[0;31m         \u001b[0mlines\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_get_lines\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdata\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[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1648\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mline\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mlines\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1649\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0madd_line\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mline\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/matplotlib/axes/_base.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m    214\u001b[0m                 \u001b[0mthis\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\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    215\u001b[0m                 \u001b[0margs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\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--> 216\u001b[0;31m             \u001b[0;32myield\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_plot_args\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mthis\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    217\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    218\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mget_next_color\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\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/usr/local/lib/python3.7/dist-packages/matplotlib/axes/_base.py\u001b[0m in \u001b[0;36m_plot_args\u001b[0;34m(self, tup, kwargs)\u001b[0m\n\u001b[1;32m    330\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtup\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m2\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    331\u001b[0m             \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_check_1d\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtup\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\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--> 332\u001b[0;31m             \u001b[0my\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_check_1d\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtup\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[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    333\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    334\u001b[0m             \u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mindex_of\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtup\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[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/matplotlib/cbook/__init__.py\u001b[0m in \u001b[0;36m_check_1d\u001b[0;34m(x)\u001b[0m\n\u001b[1;32m   1324\u001b[0m     '''\n\u001b[1;32m   1325\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mhasattr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'shape'\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\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[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1326\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0matleast_1d\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\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   1327\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   1328\u001b[0m         \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m<__array_function__ internals>\u001b[0m in \u001b[0;36matleast_1d\u001b[0;34m(*args, **kwargs)\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/numpy/core/shape_base.py\u001b[0m in \u001b[0;36matleast_1d\u001b[0;34m(*arys)\u001b[0m\n\u001b[1;32m     63\u001b[0m     \u001b[0mres\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     64\u001b[0m     \u001b[0;32mfor\u001b[0m \u001b[0mary\u001b[0m \u001b[0;32min\u001b[0m \u001b[0marys\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 65\u001b[0;31m         \u001b[0mary\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0masanyarray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mary\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     66\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mary\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mndim\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     67\u001b[0m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mary\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mreshape\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/_tensor.py\u001b[0m in \u001b[0;36m__array__\u001b[0;34m(self, dtype)\u001b[0m\n\u001b[1;32m    755\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mhandle_torch_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mTensor\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__array__\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    756\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mdtype\u001b[0m \u001b[0;32mis\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;32m--> 757\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnumpy\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    758\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    759\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnumpy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mastype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcopy\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mTypeError\u001b[0m: can't convert cuda:0 device type tensor to numpy. Use Tensor.cpu() to copy the tensor to host memory first."
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print('Testing...')\n",
        "\n",
        "def test_model():\n",
        "    \n",
        "    critic_model.eval()\n",
        "    actor_model.eval()\n",
        "    test_rew_list = []\n",
        "    for ep in range(100):\n",
        "        test_total_rew = 0\n",
        "        done = False\n",
        "        state = env.reset()\n",
        "        while not done:\n",
        "            with torch.no_grad():\n",
        "                state = torch.FloatTensor(state).cpu().to(device)\n",
        "                dist_param = actor_model(state)\n",
        "                dist = Normal(dist_param[0],dist_param[1])\n",
        "                action = dist.sample().cpu().numpy().reshape(1)\n",
        "            new_state, rew, done,_ = env.step(action)\n",
        "            test_total_rew+=rew\n",
        "            state = new_state\n",
        "        test_rew_list.append(test_total_rew)\n",
        "    return test_rew_list"
      ],
      "metadata": {
        "id": "EUpbf6Oec8Gs",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2506061b-6177-4b03-e6ea-a5350aee8374"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Testing...\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_rew = test_model()\n",
        "\n",
        "fig2,ax2 = plt.subplots()\n",
        "x2 = np.arange(len(test_rew))\n",
        "ax2.plot(x2,test_rew)\n",
        "ax2.set_xlabel('episodes')\n",
        "ax2.set_ylabel('rewards')\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "7MV6FmHlfkP7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        },
        "outputId": "8b991137-699f-4fae-fe0f-f3a8be6f8f3c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}