Project_Navigation_Banana_Collector_RL_Agent.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from unityagents import UnityEnvironment\n",
    "import numpy as np\n",
    "import torch\n",
    "import os\n",
    "\n",
    "import time\n",
    "from collections import namedtuple\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from DeepQAgent.Agent import Agent as DQNAgent\n",
    "from DeepQAgent.Agent_Priority import Agent as DQN_Prioritized_Agent\n",
    "from DeepQAgent.Agent_Priority import ExpBuffer\n",
    "\n",
    "from DeepQAgent.QNetwork import QNetwork"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_ver = 8 # Number of different Model Architectures after appropriate performance\n",
    "trial_ver = 5 # ver. 5 -> Avg Reward of 14.0 | ver. 4 -> avg Reward of 13.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def checkpoint(agent, num_checkpoint, checkpoint_path=\"./Model_Checkpoint/\",\n",
    "               model_ver=model_ver, trial_ver=trial_ver\n",
    "              ):\n",
    "    \n",
    "    # check if the Checkpoint path exists or not - if not then create it\n",
    "    if not os.path.isdir(checkpoint_path):\n",
    "        print(\"Creating Checkpoint path - Relative Path : {}\".format(checkpoint_path))\n",
    "        os.mkdir(checkpoint_path)\n",
    "    \n",
    "    # save the Trained Agent -> Local Q Network\n",
    "    checkpoint_QLocal  = {\n",
    "        'state_size': agent.state_size, 'action_size': agent.action_size, 'seed': agent.seed, \"l1_n_units\": 256,\n",
    "        \"l2_n_units\": 1024, \"l3_n_units\": 1024, \"l4_n_units\": 64,\n",
    "        'state_dict' : agent.qnetwork_local.state_dict()\n",
    "    }\n",
    "\n",
    "    torch.save( checkpoint_QLocal, checkpoint_path+\"checkpoint_{}_QLocal_{}_{}.pth\".format(num_checkpoint,\n",
    "                                                                                           model_ver, trial_ver\n",
    "                                                                                          )\n",
    "              )\n",
    "\n",
    "    # save the Trained Agent -> Target QNetwork\n",
    "    checkpoint_QTarget  = {        \n",
    "        'state_size': agent.state_size, 'action_size': agent.action_size, 'seed': agent.seed, \"l1_n_units\": 256,\n",
    "        \"l2_n_units\": 1024, \"l3_n_units\": 1024, \"l4_n_units\": 64,\n",
    "        'state_dict' : agent.qnetwork_target.state_dict()\n",
    "    }\n",
    "\n",
    "    torch.save( checkpoint_QTarget, checkpoint_path+\"checkpoint_{}_QTarget_{}_{}.pth\".format(num_checkpoint,\n",
    "                                                                                             model_ver, trial_ver\n",
    "                                                                                            )\n",
    "              )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:unityagents:\n",
      "'Academy' started successfully!\n",
      "Unity Academy name: Academy\n",
      "        Number of Brains: 1\n",
      "        Number of External Brains : 1\n",
      "        Lesson number : 0\n",
      "        Reset Parameters :\n",
      "\t\t\n",
      "Unity brain name: BananaBrain\n",
      "        Number of Visual Observations (per agent): 0\n",
      "        Vector Observation space type: continuous\n",
      "        Vector Observation space size (per agent): 37\n",
      "        Number of stacked Vector Observation: 1\n",
      "        Vector Action space type: discrete\n",
      "        Vector Action space size (per agent): 4\n",
      "        Vector Action descriptions: , , , \n"
     ]
    }
   ],
   "source": [
    "# Load the Environment\n",
    "env = UnityEnvironment(file_name=\"./Navigation_Environment/Banana.exe\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "brain_name = env.brain_names[0]\n",
    "brain = env.brains[brain_name]\n",
    "\n",
    "action_size = brain.vector_action_space_size       # get the size of the action space\n",
    "env_info = env.reset(train_mode=True)[brain_name] # reset the environment\n",
    "state = env_info.vector_observations[0]            # get the current state\n",
    "state_size = len(state)\n",
    "score = 0                                          # initialize the score\n",
    "seed = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "IS_PRIORITIZED_REPLAY = False #True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# total number of Episodes for Training Agent\n",
    "num_episodes = 10000\n",
    "\n",
    "# Define the Agent\n",
    "\n",
    "agent = None\n",
    "\n",
    "if IS_PRIORITIZED_REPLAY:\n",
    "    # Agent instantiation with Prioritized Experience Replay\n",
    "    agent = DQN_Prioritized_Agent( state_size, action_size, seed,\n",
    "                                   torch.device('cuda:0' if torch.cuda.is_available() else 'cpu:0')\n",
    "                                 )\n",
    "else:\n",
    "    # Agent instantiation without Prioritized Experience Replay\n",
    "    agent = DQNAgent( state_size, action_size, seed )\n",
    "\n",
    "epsilon_thresh = 0.1 # 0.1\n",
    "epsilon_decay = 0.99 # 0.999995 # find a decay factor so that epsilon decays each timestep to close to 0 in One episode\n",
    "episodic_epsilon_decay = 0.985\n",
    "epi_epsilon = 1/(episodic_epsilon_decay) # The first Episode should start from 1.0 -> we'll be decaying over Episodes too\n",
    "\n",
    "episode_score = []\n",
    "inspect_episodes = 100\n",
    "avg_score_per_inspect_episodes = [] # store the average score per elapsed 'inspect-episodes'\n",
    "num_checkpoint = 1\n",
    "inspect_score_thresh = 13.0 # Threshold of Average score over 'inspect_episodes' after which terminate simulation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load from check point if any\n",
    "# agent.qnet_local.load_state_dict( torch.load('checkpoint_1_QLocal.pth')['state_dict'] )\n",
    "# agent.qnet_target.load_state_dict( torch.load('checkpoint_1_QTarget.pth')['state_dict'] )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Episode: 1 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 0 Yellow: 0 > Elapsed: 3.13sec            \n",
      "Episode: 2 ~ Tstep: 301| Score: -1.0 Avg Score : 0 | Blue: 1 Yellow: 0 > Elapsed: 1.7sec            \n",
      "Episode: 3 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 0 Yellow: 1 > Elapsed: 1.76sec            \n",
      "Episode: 4 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 0 Yellow: 1 > Elapsed: 1.71sec            \n",
      "Episode: 5 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 0 Yellow: 0 > Elapsed: 1.87sec            \n",
      "Episode: 6 ~ Tstep: 301| Score: -1.0 Avg Score : 0 | Blue: 1 Yellow: 0 > Elapsed: 1.76sec           \n",
      "Episode: 7 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 1 Yellow: 1 > Elapsed: 1.75sec            \n",
      "Episode: 8 ~ Tstep: 301| Score: -1.0 Avg Score : 0 | Blue: 2 Yellow: 1 > Elapsed: 1.76sec           \n",
      "Episode: 9 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 0 Yellow: 0 > Elapsed: 1.78sec            \n",
      "Episode: 10 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 1 Yellow: 1 > Elapsed: 1.69sec           \n",
      "Episode: 11 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 1 Yellow: 2 > Elapsed: 1.77sec           \n",
      "Episode: 12 ~ Tstep: 301| Score: 4.0 Avg Score : 0 | Blue: 0 Yellow: 4 > Elapsed: 1.76sec           \n",
      "Episode: 13 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 2 Yellow: 2 > Elapsed: 1.74sec           \n",
      "Episode: 14 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 0 Yellow: 1 > Elapsed: 1.75sec           \n",
      "Episode: 15 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 2 Yellow: 2 > Elapsed: 1.72sec           \n",
      "Episode: 16 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 1 Yellow: 1 > Elapsed: 1.78sec           \n",
      "Episode: 17 ~ Tstep: 301| Score: 3.0 Avg Score : 0 | Blue: 0 Yellow: 3 > Elapsed: 1.74sec           \n",
      "Episode: 18 ~ Tstep: 301| Score: 3.0 Avg Score : 0 | Blue: 1 Yellow: 4 > Elapsed: 1.74sec           \n",
      "Episode: 19 ~ Tstep: 301| Score: 3.0 Avg Score : 0 | Blue: 1 Yellow: 4 > Elapsed: 1.73sec           \n",
      "Episode: 20 ~ Tstep: 301| Score: 3.0 Avg Score : 0 | Blue: 0 Yellow: 3 > Elapsed: 1.74sec           \n",
      "Episode: 21 ~ Tstep: 301| Score: 8.0 Avg Score : 0 | Blue: 0 Yellow: 8 > Elapsed: 1.79sec           \n",
      "Episode: 22 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 2 Yellow: 2 > Elapsed: 1.72sec           \n",
      "Episode: 23 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 0 Yellow: 0 > Elapsed: 1.77sec           \n",
      "Episode: 24 ~ Tstep: 301| Score: 6.0 Avg Score : 0 | Blue: 0 Yellow: 6 > Elapsed: 1.76sec           \n",
      "Episode: 25 ~ Tstep: 301| Score: 2.0 Avg Score : 0 | Blue: 1 Yellow: 3 > Elapsed: 1.75sec           \n",
      "Episode: 26 ~ Tstep: 301| Score: 2.0 Avg Score : 0 | Blue: 2 Yellow: 4 > Elapsed: 1.76sec           \n",
      "Episode: 27 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 1 Yellow: 1 > Elapsed: 1.71sec           \n",
      "Episode: 28 ~ Tstep: 301| Score: 3.0 Avg Score : 0 | Blue: 1 Yellow: 4 > Elapsed: 1.72sec           \n",
      "Episode: 29 ~ Tstep: 301| Score: 4.0 Avg Score : 0 | Blue: 0 Yellow: 4 > Elapsed: 1.75sec           \n",
      "Episode: 30 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 1 Yellow: 2 > Elapsed: 1.75sec           \n",
      "Episode: 31 ~ Tstep: 301| Score: 3.0 Avg Score : 0 | Blue: 0 Yellow: 3 > Elapsed: 1.73sec           \n",
      "Episode: 32 ~ Tstep: 301| Score: 6.0 Avg Score : 0 | Blue: 1 Yellow: 7 > Elapsed: 1.77sec           \n",
      "Episode: 33 ~ Tstep: 301| Score: 8.0 Avg Score : 0 | Blue: 0 Yellow: 8 > Elapsed: 1.72sec           \n",
      "Episode: 34 ~ Tstep: 301| Score: 4.0 Avg Score : 0 | Blue: 0 Yellow: 4 > Elapsed: 1.75sec           \n",
      "Episode: 35 ~ Tstep: 301| Score: 4.0 Avg Score : 0 | Blue: 0 Yellow: 4 > Elapsed: 1.77sec           \n",
      "Episode: 36 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 2 Yellow: 3 > Elapsed: 1.76sec           \n",
      "Episode: 37 ~ Tstep: 301| Score: 3.0 Avg Score : 0 | Blue: 1 Yellow: 4 > Elapsed: 1.72sec           \n",
      "Episode: 38 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 1 Yellow: 2 > Elapsed: 1.83sec           \n",
      "Episode: 39 ~ Tstep: 301| Score: 2.0 Avg Score : 0 | Blue: 0 Yellow: 2 > Elapsed: 1.7sec            \n",
      "Episode: 40 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 0 Yellow: 1 > Elapsed: 1.73sec           \n",
      "Episode: 41 ~ Tstep: 301| Score: 6.0 Avg Score : 0 | Blue: 0 Yellow: 6 > Elapsed: 1.73sec           \n",
      "Episode: 42 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 0 Yellow: 1 > Elapsed: 1.77sec           \n",
      "Episode: 43 ~ Tstep: 301| Score: 4.0 Avg Score : 0 | Blue: 1 Yellow: 5 > Elapsed: 1.73sec           \n",
      "Episode: 44 ~ Tstep: 301| Score: 2.0 Avg Score : 0 | Blue: 0 Yellow: 2 > Elapsed: 1.71sec           \n",
      "Episode: 45 ~ Tstep: 301| Score: 2.0 Avg Score : 0 | Blue: 0 Yellow: 2 > Elapsed: 1.75sec           \n",
      "Episode: 46 ~ Tstep: 301| Score: 3.0 Avg Score : 0 | Blue: 0 Yellow: 3 > Elapsed: 1.75sec           \n",
      "Episode: 47 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 2 Yellow: 3 > Elapsed: 1.74sec           \n",
      "Episode: 48 ~ Tstep: 301| Score: 6.0 Avg Score : 0 | Blue: 0 Yellow: 6 > Elapsed: 1.79sec           \n",
      "Episode: 49 ~ Tstep: 301| Score: 7.0 Avg Score : 0 | Blue: 0 Yellow: 7 > Elapsed: 1.73sec           \n",
      "Episode: 50 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 1 Yellow: 2 > Elapsed: 1.73sec           \n",
      "Episode: 51 ~ Tstep: 301| Score: 3.0 Avg Score : 0 | Blue: 0 Yellow: 3 > Elapsed: 1.74sec           \n",
      "Episode: 52 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 0 Yellow: 0 > Elapsed: 1.72sec           \n",
      "Episode: 53 ~ Tstep: 301| Score: 4.0 Avg Score : 0 | Blue: 0 Yellow: 4 > Elapsed: 1.83sec           \n",
      "Episode: 54 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 0 Yellow: 0 > Elapsed: 1.72sec           \n",
      "Episode: 55 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 2 Yellow: 2 > Elapsed: 1.74sec           \n",
      "Episode: 56 ~ Tstep: 301| Score: 3.0 Avg Score : 0 | Blue: 0 Yellow: 3 > Elapsed: 1.73sec           \n",
      "Episode: 57 ~ Tstep: 301| Score: 5.0 Avg Score : 0 | Blue: 0 Yellow: 5 > Elapsed: 1.76sec           \n",
      "Episode: 58 ~ Tstep: 301| Score: 3.0 Avg Score : 0 | Blue: 1 Yellow: 4 > Elapsed: 1.75sec           \n",
      "Episode: 59 ~ Tstep: 301| Score: 4.0 Avg Score : 0 | Blue: 1 Yellow: 5 > Elapsed: 1.83sec           \n",
      "Episode: 60 ~ Tstep: 301| Score: 6.0 Avg Score : 0 | Blue: 0 Yellow: 6 > Elapsed: 1.7sec            \n",
      "Episode: 61 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 1 Yellow: 2 > Elapsed: 1.78sec           \n",
      "Episode: 62 ~ Tstep: 301| Score: 2.0 Avg Score : 0 | Blue: 3 Yellow: 5 > Elapsed: 1.73sec           \n",
      "Episode: 63 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 0 Yellow: 1 > Elapsed: 1.74sec           \n",
      "Episode: 64 ~ Tstep: 301| Score: 2.0 Avg Score : 0 | Blue: 0 Yellow: 2 > Elapsed: 1.72sec           \n",
      "Episode: 65 ~ Tstep: 301| Score: 11.0 Avg Score : 0 | Blue: 0 Yellow: 11 > Elapsed: 1.73sec         \n",
      "Episode: 66 ~ Tstep: 301| Score: 4.0 Avg Score : 0 | Blue: 2 Yellow: 6 > Elapsed: 1.77sec           \n",
      "Episode: 67 ~ Tstep: 301| Score: 2.0 Avg Score : 0 | Blue: 0 Yellow: 2 > Elapsed: 1.73sec           \n",
      "Episode: 68 ~ Tstep: 301| Score: -1.0 Avg Score : 0 | Blue: 2 Yellow: 1 > Elapsed: 1.77sec          \n",
      "Episode: 69 ~ Tstep: 301| Score: -3.0 Avg Score : 0 | Blue: 3 Yellow: 0 > Elapsed: 1.76sec          \n",
      "Episode: 70 ~ Tstep: 301| Score: 3.0 Avg Score : 0 | Blue: 2 Yellow: 5 > Elapsed: 1.76sec           \n",
      "Episode: 71 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 1 Yellow: 2 > Elapsed: 1.71sec           \n",
      "Episode: 72 ~ Tstep: 301| Score: 4.0 Avg Score : 0 | Blue: 0 Yellow: 4 > Elapsed: 1.77sec           \n",
      "Episode: 73 ~ Tstep: 301| Score: 7.0 Avg Score : 0 | Blue: 0 Yellow: 7 > Elapsed: 1.73sec           \n",
      "Episode: 74 ~ Tstep: 301| Score: 3.0 Avg Score : 0 | Blue: 1 Yellow: 4 > Elapsed: 1.76sec           \n",
      "Episode: 75 ~ Tstep: 301| Score: -2.0 Avg Score : 0 | Blue: 2 Yellow: 0 > Elapsed: 1.74sec          \n",
      "Episode: 76 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 1 Yellow: 2 > Elapsed: 1.73sec           \n",
      "Episode: 77 ~ Tstep: 301| Score: 8.0 Avg Score : 0 | Blue: 2 Yellow: 10 > Elapsed: 1.78sec          \n",
      "Episode: 78 ~ Tstep: 301| Score: 5.0 Avg Score : 0 | Blue: 0 Yellow: 5 > Elapsed: 1.75sec           \n",
      "Episode: 79 ~ Tstep: 301| Score: 3.0 Avg Score : 0 | Blue: 2 Yellow: 5 > Elapsed: 1.98sec           \n",
      "Episode: 80 ~ Tstep: 301| Score: -3.0 Avg Score : 0 | Blue: 4 Yellow: 1 > Elapsed: 1.78sec          \n",
      "Episode: 81 ~ Tstep: 301| Score: 11.0 Avg Score : 0 | Blue: 0 Yellow: 11 > Elapsed: 1.75sec         \n",
      "Episode: 82 ~ Tstep: 301| Score: 3.0 Avg Score : 0 | Blue: 4 Yellow: 7 > Elapsed: 1.77sec           \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Episode: 83 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 2 Yellow: 2 > Elapsed: 1.79sec           \n",
      "Episode: 84 ~ Tstep: 301| Score: 4.0 Avg Score : 0 | Blue: 2 Yellow: 6 > Elapsed: 1.76sec           \n",
      "Episode: 85 ~ Tstep: 301| Score: 0.0 Avg Score : 0 | Blue: 1 Yellow: 1 > Elapsed: 1.72sec           \n",
      "Episode: 86 ~ Tstep: 301| Score: 9.0 Avg Score : 0 | Blue: 1 Yellow: 10 > Elapsed: 1.71sec          \n",
      "Episode: 87 ~ Tstep: 301| Score: 6.0 Avg Score : 0 | Blue: 0 Yellow: 6 > Elapsed: 1.71sec           \n",
      "Episode: 88 ~ Tstep: 301| Score: 12.0 Avg Score : 0 | Blue: 1 Yellow: 13 > Elapsed: 1.76sec         \n",
      "Episode: 89 ~ Tstep: 301| Score: 14.0 Avg Score : 0 | Blue: 1 Yellow: 15 > Elapsed: 1.74sec         \n",
      "Episode: 90 ~ Tstep: 301| Score: 4.0 Avg Score : 0 | Blue: 0 Yellow: 4 > Elapsed: 1.72sec           \n",
      "Episode: 91 ~ Tstep: 301| Score: 5.0 Avg Score : 0 | Blue: 2 Yellow: 7 > Elapsed: 1.89sec           \n",
      "Episode: 92 ~ Tstep: 301| Score: 5.0 Avg Score : 0 | Blue: 1 Yellow: 6 > Elapsed: 1.76sec           \n",
      "Episode: 93 ~ Tstep: 301| Score: 5.0 Avg Score : 0 | Blue: 3 Yellow: 8 > Elapsed: 1.78sec           \n",
      "Episode: 94 ~ Tstep: 301| Score: 11.0 Avg Score : 0 | Blue: 0 Yellow: 11 > Elapsed: 2.1sec          \n",
      "Episode: 95 ~ Tstep: 301| Score: 7.0 Avg Score : 0 | Blue: 0 Yellow: 7 > Elapsed: 1.73sec           \n",
      "Episode: 96 ~ Tstep: 301| Score: 11.0 Avg Score : 0 | Blue: 1 Yellow: 12 > Elapsed: 1.76sec         \n",
      "Episode: 97 ~ Tstep: 301| Score: 5.0 Avg Score : 0 | Blue: 0 Yellow: 5 > Elapsed: 1.75sec           \n",
      "Episode: 98 ~ Tstep: 301| Score: 4.0 Avg Score : 0 | Blue: 1 Yellow: 5 > Elapsed: 1.74sec           \n",
      "Episode: 99 ~ Tstep: 301| Score: 1.0 Avg Score : 0 | Blue: 2 Yellow: 3 > Elapsed: 1.76sec           \n",
      "Episode: 100 ~ Tstep: 301| Score: 11.0 Avg Score : 0 | Blue: 2 Yellow: 13 > Elapsed: 1.74sec        \n",
      "Episode: 101 ~ Tstep: 301| Score: 4.0 Avg Score : 3.18 | Blue: 2 Yellow: 6 > Elapsed: 1.72sec       \n",
      "Episode: 102 ~ Tstep: 301| Score: 4.0 Avg Score : 3.23 | Blue: 3 Yellow: 7 > Elapsed: 1.76sec       \n",
      "Episode: 103 ~ Tstep: 301| Score: 10.0 Avg Score : 3.32 | Blue: 0 Yellow: 10 > Elapsed: 1.79sec     \n",
      "Episode: 104 ~ Tstep: 301| Score: 9.0 Avg Score : 3.4 | Blue: 0 Yellow: 9 > Elapsed: 1.74sec        \n",
      "Episode: 105 ~ Tstep: 301| Score: 9.0 Avg Score : 3.49 | Blue: 1 Yellow: 10 > Elapsed: 1.74sec      \n",
      "Episode: 106 ~ Tstep: 301| Score: 7.0 Avg Score : 3.57 | Blue: 1 Yellow: 8 > Elapsed: 1.77sec       \n",
      "Episode: 107 ~ Tstep: 301| Score: 4.0 Avg Score : 3.61 | Blue: 0 Yellow: 4 > Elapsed: 1.73sec       \n",
      "Episode: 108 ~ Tstep: 301| Score: 11.0 Avg Score : 3.73 | Blue: 0 Yellow: 11 > Elapsed: 1.75sec     \n",
      "Episode: 109 ~ Tstep: 301| Score: 6.0 Avg Score : 3.79 | Blue: 1 Yellow: 7 > Elapsed: 1.76sec       \n",
      "Episode: 110 ~ Tstep: 301| Score: 5.0 Avg Score : 3.84 | Blue: 1 Yellow: 6 > Elapsed: 1.7sec        \n",
      "Episode: 111 ~ Tstep: 301| Score: 14.0 Avg Score : 3.97 | Blue: 1 Yellow: 15 > Elapsed: 1.75sec     \n",
      "Episode: 112 ~ Tstep: 301| Score: 8.0 Avg Score : 4.01 | Blue: 1 Yellow: 9 > Elapsed: 1.73sec       \n",
      "Episode: 113 ~ Tstep: 301| Score: 13.0 Avg Score : 4.14 | Blue: 1 Yellow: 14 > Elapsed: 1.77sec     \n",
      "Episode: 114 ~ Tstep: 301| Score: 9.0 Avg Score : 4.22 | Blue: 0 Yellow: 9 > Elapsed: 1.75sec       \n",
      "Episode: 115 ~ Tstep: 301| Score: 7.0 Avg Score : 4.29 | Blue: 0 Yellow: 7 > Elapsed: 1.78sec       \n",
      "Episode: 116 ~ Tstep: 301| Score: -1.0 Avg Score : 4.28 | Blue: 1 Yellow: 0 > Elapsed: 1.71sec      \n",
      "Episode: 117 ~ Tstep: 301| Score: 7.0 Avg Score : 4.32 | Blue: 1 Yellow: 8 > Elapsed: 1.78sec       \n",
      "Episode: 118 ~ Tstep: 301| Score: 11.0 Avg Score : 4.4 | Blue: 0 Yellow: 11 > Elapsed: 1.76sec      \n",
      "Episode: 119 ~ Tstep: 301| Score: 4.0 Avg Score : 4.41 | Blue: 3 Yellow: 7 > Elapsed: 1.72sec       \n",
      "Episode: 120 ~ Tstep: 301| Score: 11.0 Avg Score : 4.49 | Blue: 0 Yellow: 11 > Elapsed: 1.73sec     \n",
      "Episode: 121 ~ Tstep: 301| Score: 1.0 Avg Score : 4.42 | Blue: 1 Yellow: 2 > Elapsed: 1.75sec       \n",
      "Episode: 122 ~ Tstep: 301| Score: 15.0 Avg Score : 4.57 | Blue: 1 Yellow: 16 > Elapsed: 1.81sec     \n",
      "Episode: 123 ~ Tstep: 301| Score: 6.0 Avg Score : 4.63 | Blue: 1 Yellow: 7 > Elapsed: 1.74sec       \n",
      "Episode: 124 ~ Tstep: 301| Score: 7.0 Avg Score : 4.64 | Blue: 0 Yellow: 7 > Elapsed: 1.71sec       \n",
      "Episode: 125 ~ Tstep: 301| Score: 1.0 Avg Score : 4.63 | Blue: 0 Yellow: 1 > Elapsed: 1.78sec       \n",
      "Episode: 126 ~ Tstep: 301| Score: 13.0 Avg Score : 4.74 | Blue: 0 Yellow: 13 > Elapsed: 1.75sec     \n",
      "Episode: 127 ~ Tstep: 301| Score: 3.0 Avg Score : 4.77 | Blue: 0 Yellow: 3 > Elapsed: 1.73sec       \n",
      "Episode: 128 ~ Tstep: 301| Score: 5.0 Avg Score : 4.79 | Blue: 0 Yellow: 5 > Elapsed: 1.75sec       \n",
      "Episode: 129 ~ Tstep: 301| Score: 12.0 Avg Score : 4.87 | Blue: 0 Yellow: 12 > Elapsed: 1.76sec     \n",
      "Episode: 130 ~ Tstep: 301| Score: 8.0 Avg Score : 4.94 | Blue: 0 Yellow: 8 > Elapsed: 1.75sec       \n",
      "Episode: 131 ~ Tstep: 301| Score: 12.0 Avg Score : 5.03 | Blue: 0 Yellow: 12 > Elapsed: 1.74sec     \n",
      "Episode: 132 ~ Tstep: 301| Score: 7.0 Avg Score : 5.04 | Blue: 0 Yellow: 7 > Elapsed: 1.71sec       \n",
      "Episode: 133 ~ Tstep: 301| Score: 5.0 Avg Score : 5.01 | Blue: 0 Yellow: 5 > Elapsed: 1.79sec       \n",
      "Episode: 134 ~ Tstep: 301| Score: 16.0 Avg Score : 5.13 | Blue: 1 Yellow: 17 > Elapsed: 1.74sec     \n",
      "Episode: 135 ~ Tstep: 301| Score: 13.0 Avg Score : 5.22 | Blue: 0 Yellow: 13 > Elapsed: 1.77sec     \n",
      "Episode: 136 ~ Tstep: 301| Score: 7.0 Avg Score : 5.28 | Blue: 1 Yellow: 8 > Elapsed: 1.77sec       \n",
      "Episode: 137 ~ Tstep: 301| Score: 9.0 Avg Score : 5.34 | Blue: 1 Yellow: 10 > Elapsed: 1.75sec      \n",
      "Episode: 138 ~ Tstep: 301| Score: 4.0 Avg Score : 5.37 | Blue: 1 Yellow: 5 > Elapsed: 1.74sec       \n",
      "Episode: 139 ~ Tstep: 301| Score: 18.0 Avg Score : 5.53 | Blue: 0 Yellow: 18 > Elapsed: 1.76sec     \n",
      "Episode: 140 ~ Tstep: 301| Score: 9.0 Avg Score : 5.61 | Blue: 0 Yellow: 9 > Elapsed: 1.72sec       \n",
      "Episode: 141 ~ Tstep: 301| Score: 10.0 Avg Score : 5.65 | Blue: 0 Yellow: 10 > Elapsed: 1.75sec     \n",
      "Episode: 142 ~ Tstep: 301| Score: 3.0 Avg Score : 5.67 | Blue: 1 Yellow: 4 > Elapsed: 1.74sec       \n",
      "Episode: 143 ~ Tstep: 301| Score: 11.0 Avg Score : 5.74 | Blue: 0 Yellow: 11 > Elapsed: 1.8sec      \n",
      "Episode: 144 ~ Tstep: 301| Score: 11.0 Avg Score : 5.83 | Blue: 1 Yellow: 12 > Elapsed: 1.71sec     \n",
      "Episode: 145 ~ Tstep: 301| Score: 5.0 Avg Score : 5.86 | Blue: 3 Yellow: 7 > Elapsed: 1.71sec       \n",
      "Episode: 146 ~ Tstep: 301| Score: 7.0 Avg Score : 5.9 | Blue: 0 Yellow: 7 > Elapsed: 1.73sec        \n",
      "Episode: 147 ~ Tstep: 301| Score: 9.0 Avg Score : 5.98 | Blue: 0 Yellow: 9 > Elapsed: 1.76sec       \n",
      "Episode: 148 ~ Tstep: 301| Score: 11.0 Avg Score : 6.03 | Blue: 1 Yellow: 12 > Elapsed: 1.75sec     \n",
      "Episode: 149 ~ Tstep: 301| Score: 8.0 Avg Score : 6.04 | Blue: 0 Yellow: 8 > Elapsed: 1.75sec       \n",
      "Episode: 150 ~ Tstep: 301| Score: 5.0 Avg Score : 6.08 | Blue: 0 Yellow: 5 > Elapsed: 1.77sec       \n",
      "Episode: 151 ~ Tstep: 301| Score: 14.0 Avg Score : 6.19 | Blue: 1 Yellow: 15 > Elapsed: 1.75sec     \n",
      "Episode: 152 ~ Tstep: 301| Score: 5.0 Avg Score : 6.24 | Blue: 0 Yellow: 5 > Elapsed: 1.71sec       \n",
      "Episode: 153 ~ Tstep: 301| Score: 6.0 Avg Score : 6.26 | Blue: 1 Yellow: 7 > Elapsed: 1.75sec       \n",
      "Episode: 154 ~ Tstep: 301| Score: 10.0 Avg Score : 6.36 | Blue: 0 Yellow: 10 > Elapsed: 1.77sec     \n",
      "Episode: 155 ~ Tstep: 301| Score: 15.0 Avg Score : 6.51 | Blue: 1 Yellow: 16 > Elapsed: 1.78sec     \n",
      "Episode: 156 ~ Tstep: 301| Score: 11.0 Avg Score : 6.59 | Blue: 0 Yellow: 11 > Elapsed: 1.75sec     \n",
      "Episode: 157 ~ Tstep: 301| Score: 7.0 Avg Score : 6.61 | Blue: 0 Yellow: 7 > Elapsed: 1.75sec       \n",
      "Episode: 158 ~ Tstep: 301| Score: 11.0 Avg Score : 6.69 | Blue: 1 Yellow: 12 > Elapsed: 1.76sec     \n",
      "Episode: 159 ~ Tstep: 301| Score: 8.0 Avg Score : 6.73 | Blue: 2 Yellow: 10 > Elapsed: 1.75sec      \n",
      "Episode: 160 ~ Tstep: 301| Score: 8.0 Avg Score : 6.75 | Blue: 1 Yellow: 9 > Elapsed: 1.69sec       \n",
      "Episode: 161 ~ Tstep: 301| Score: 10.0 Avg Score : 6.84 | Blue: 1 Yellow: 11 > Elapsed: 1.74sec     \n",
      "Episode: 162 ~ Tstep: 301| Score: 2.0 Avg Score : 6.84 | Blue: 1 Yellow: 3 > Elapsed: 1.73sec       \n",
      "Episode: 163 ~ Tstep: 301| Score: 15.0 Avg Score : 6.98 | Blue: 1 Yellow: 16 > Elapsed: 1.73sec     \n",
      "Episode: 164 ~ Tstep: 301| Score: 13.0 Avg Score : 7.09 | Blue: 0 Yellow: 13 > Elapsed: 1.75sec     \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Episode: 165 ~ Tstep: 301| Score: 0.0 Avg Score : 6.98 | Blue: 0 Yellow: 0 > Elapsed: 1.72sec       \n",
      "Episode: 166 ~ Tstep: 301| Score: 10.0 Avg Score : 7.04 | Blue: 2 Yellow: 12 > Elapsed: 1.74sec     \n",
      "Episode: 167 ~ Tstep: 301| Score: 14.0 Avg Score : 7.16 | Blue: 0 Yellow: 14 > Elapsed: 1.76sec     \n",
      "Episode: 168 ~ Tstep: 301| Score: 14.0 Avg Score : 7.31 | Blue: 0 Yellow: 14 > Elapsed: 1.76sec     \n",
      "Episode: 169 ~ Tstep: 301| Score: 8.0 Avg Score : 7.42 | Blue: 0 Yellow: 8 > Elapsed: 1.91sec       \n",
      "Episode: 170 ~ Tstep: 301| Score: 10.0 Avg Score : 7.49 | Blue: 0 Yellow: 10 > Elapsed: 1.78sec     \n",
      "Episode: 171 ~ Tstep: 301| Score: 8.0 Avg Score : 7.56 | Blue: 0 Yellow: 8 > Elapsed: 1.76sec       \n",
      "Episode: 172 ~ Tstep: 301| Score: 3.0 Avg Score : 7.55 | Blue: 0 Yellow: 3 > Elapsed: 1.71sec       \n",
      "Episode: 173 ~ Tstep: 301| Score: 9.0 Avg Score : 7.57 | Blue: 0 Yellow: 9 > Elapsed: 1.75sec       \n",
      "Episode: 174 ~ Tstep: 301| Score: 12.0 Avg Score : 7.66 | Blue: 0 Yellow: 12 > Elapsed: 1.76sec     \n",
      "Episode: 175 ~ Tstep: 301| Score: 21.0 Avg Score : 7.89 | Blue: 0 Yellow: 21 > Elapsed: 1.75sec     \n",
      "Episode: 176 ~ Tstep: 301| Score: 7.0 Avg Score : 7.95 | Blue: 0 Yellow: 7 > Elapsed: 1.75sec       \n",
      "Episode: 177 ~ Tstep: 301| Score: 6.0 Avg Score : 7.93 | Blue: 0 Yellow: 6 > Elapsed: 1.72sec       \n",
      "Episode: 178 ~ Tstep: 301| Score: 13.0 Avg Score : 8.01 | Blue: 0 Yellow: 13 > Elapsed: 1.82sec     \n",
      "Episode: 179 ~ Tstep: 301| Score: 11.0 Avg Score : 8.09 | Blue: 0 Yellow: 11 > Elapsed: 1.74sec     \n",
      "Episode: 180 ~ Tstep: 301| Score: 15.0 Avg Score : 8.27 | Blue: 0 Yellow: 15 > Elapsed: 1.74sec     \n",
      "Episode: 181 ~ Tstep: 301| Score: 6.0 Avg Score : 8.22 | Blue: 1 Yellow: 7 > Elapsed: 2.12sec       \n",
      "Episode: 182 ~ Tstep: 301| Score: 14.0 Avg Score : 8.33 | Blue: 0 Yellow: 14 > Elapsed: 1.71sec     \n",
      "Episode: 183 ~ Tstep: 301| Score: 10.0 Avg Score : 8.43 | Blue: 0 Yellow: 10 > Elapsed: 1.78sec     \n",
      "Episode: 184 ~ Tstep: 301| Score: 14.0 Avg Score : 8.53 | Blue: 0 Yellow: 14 > Elapsed: 1.79sec     \n",
      "Episode: 185 ~ Tstep: 301| Score: 4.0 Avg Score : 8.57 | Blue: 0 Yellow: 4 > Elapsed: 1.74sec       \n",
      "Episode: 186 ~ Tstep: 301| Score: 17.0 Avg Score : 8.65 | Blue: 1 Yellow: 18 > Elapsed: 1.87sec     \n",
      "Episode: 187 ~ Tstep: 301| Score: 6.0 Avg Score : 8.65 | Blue: 0 Yellow: 6 > Elapsed: 1.84sec       \n",
      "Episode: 188 ~ Tstep: 301| Score: 4.0 Avg Score : 8.57 | Blue: 0 Yellow: 4 > Elapsed: 1.92sec       \n",
      "Episode: 189 ~ Tstep: 301| Score: 15.0 Avg Score : 8.58 | Blue: 0 Yellow: 15 > Elapsed: 1.86sec     \n",
      "Episode: 190 ~ Tstep: 301| Score: 8.0 Avg Score : 8.62 | Blue: 0 Yellow: 8 > Elapsed: 1.72sec       \n",
      "Episode: 191 ~ Tstep: 301| Score: 12.0 Avg Score : 8.69 | Blue: 0 Yellow: 12 > Elapsed: 1.78sec     \n",
      "Episode: 192 ~ Tstep: 301| Score: 11.0 Avg Score : 8.75 | Blue: 1 Yellow: 12 > Elapsed: 1.69sec     \n",
      "Episode: 193 ~ Tstep: 301| Score: 10.0 Avg Score : 8.8 | Blue: 0 Yellow: 10 > Elapsed: 1.76sec      \n",
      "Episode: 194 ~ Tstep: 301| Score: 18.0 Avg Score : 8.87 | Blue: 1 Yellow: 19 > Elapsed: 1.75sec     \n",
      "Episode: 195 ~ Tstep: 301| Score: 11.0 Avg Score : 8.91 | Blue: 3 Yellow: 14 > Elapsed: 1.73sec     \n",
      "Episode: 196 ~ Tstep: 301| Score: 11.0 Avg Score : 8.91 | Blue: 1 Yellow: 11 > Elapsed: 1.92sec     \n",
      "Episode: 197 ~ Tstep: 301| Score: 16.0 Avg Score : 9.02 | Blue: 1 Yellow: 17 > Elapsed: 1.75sec     \n",
      "Episode: 198 ~ Tstep: 301| Score: 9.0 Avg Score : 9.07 | Blue: 0 Yellow: 9 > Elapsed: 1.75sec       \n",
      "Episode: 199 ~ Tstep: 301| Score: 13.0 Avg Score : 9.19 | Blue: 1 Yellow: 14 > Elapsed: 1.76sec     \n",
      "Episode: 200 ~ Tstep: 301| Score: 14.0 Avg Score : 9.22 | Blue: 0 Yellow: 14 > Elapsed: 1.76sec     \n",
      "Episode: 201 ~ Tstep: 301| Score: 13.0 Avg Score : 9.31 | Blue: 1 Yellow: 14 > Elapsed: 1.7sec      \n",
      "Episode: 202 ~ Tstep: 301| Score: 15.0 Avg Score : 9.42 | Blue: 2 Yellow: 17 > Elapsed: 1.76sec     \n",
      "Episode: 203 ~ Tstep: 301| Score: 12.0 Avg Score : 9.44 | Blue: 0 Yellow: 12 > Elapsed: 1.73sec     \n",
      "Episode: 204 ~ Tstep: 301| Score: 14.0 Avg Score : 9.49 | Blue: 1 Yellow: 15 > Elapsed: 1.78sec     \n",
      "Episode: 205 ~ Tstep: 301| Score: 5.0 Avg Score : 9.45 | Blue: 0 Yellow: 5 > Elapsed: 1.73sec       \n",
      "Episode: 206 ~ Tstep: 301| Score: 16.0 Avg Score : 9.54 | Blue: 0 Yellow: 16 > Elapsed: 1.76sec     \n",
      "Episode: 207 ~ Tstep: 301| Score: 17.0 Avg Score : 9.67 | Blue: 1 Yellow: 18 > Elapsed: 1.71sec     \n",
      "Episode: 208 ~ Tstep: 301| Score: 13.0 Avg Score : 9.69 | Blue: 0 Yellow: 13 > Elapsed: 1.73sec     \n",
      "Episode: 209 ~ Tstep: 301| Score: 13.0 Avg Score : 9.76 | Blue: 0 Yellow: 13 > Elapsed: 1.72sec     \n",
      "Episode: 210 ~ Tstep: 301| Score: 15.0 Avg Score : 9.86 | Blue: 1 Yellow: 16 > Elapsed: 1.75sec     \n",
      "Episode: 211 ~ Tstep: 301| Score: 15.0 Avg Score : 9.87 | Blue: 0 Yellow: 14 > Elapsed: 1.79sec     \n",
      "Episode: 212 ~ Tstep: 301| Score: 15.0 Avg Score : 9.94 | Blue: 0 Yellow: 15 > Elapsed: 1.72sec     \n",
      "Episode: 213 ~ Tstep: 301| Score: 14.0 Avg Score : 9.95 | Blue: 0 Yellow: 14 > Elapsed: 1.74sec     \n",
      "Episode: 214 ~ Tstep: 301| Score: 16.0 Avg Score : 10.02 | Blue: 0 Yellow: 16 > Elapsed: 1.74sec    \n",
      "Episode: 215 ~ Tstep: 301| Score: 19.0 Avg Score : 10.14 | Blue: 0 Yellow: 19 > Elapsed: 1.74sec    \n",
      "Episode: 216 ~ Tstep: 301| Score: 8.0 Avg Score : 10.23 | Blue: 0 Yellow: 8 > Elapsed: 1.78sec      \n",
      "Episode: 217 ~ Tstep: 301| Score: 10.0 Avg Score : 10.26 | Blue: 1 Yellow: 11 > Elapsed: 1.83sec    \n",
      "Episode: 218 ~ Tstep: 301| Score: 16.0 Avg Score : 10.31 | Blue: 0 Yellow: 16 > Elapsed: 1.73sec    \n",
      "Episode: 219 ~ Tstep: 301| Score: 14.0 Avg Score : 10.41 | Blue: 0 Yellow: 14 > Elapsed: 1.7sec     \n",
      "Episode: 220 ~ Tstep: 301| Score: 13.0 Avg Score : 10.43 | Blue: 1 Yellow: 14 > Elapsed: 1.71sec    \n",
      "Episode: 221 ~ Tstep: 301| Score: 12.0 Avg Score : 10.54 | Blue: 0 Yellow: 12 > Elapsed: 1.79sec    \n",
      "Episode: 222 ~ Tstep: 301| Score: 15.0 Avg Score : 10.54 | Blue: 0 Yellow: 15 > Elapsed: 1.74sec    \n",
      "Episode: 223 ~ Tstep: 301| Score: 15.0 Avg Score : 10.63 | Blue: 0 Yellow: 15 > Elapsed: 1.77sec    \n",
      "Episode: 224 ~ Tstep: 301| Score: 15.0 Avg Score : 10.71 | Blue: 1 Yellow: 16 > Elapsed: 1.76sec    \n",
      "Episode: 225 ~ Tstep: 301| Score: 9.0 Avg Score : 10.79 | Blue: 0 Yellow: 9 > Elapsed: 1.7sec       \n",
      "Episode: 226 ~ Tstep: 301| Score: 15.0 Avg Score : 10.81 | Blue: 1 Yellow: 16 > Elapsed: 1.76sec    \n",
      "Episode: 227 ~ Tstep: 301| Score: 11.0 Avg Score : 10.89 | Blue: 0 Yellow: 11 > Elapsed: 1.73sec    \n",
      "Episode: 228 ~ Tstep: 301| Score: 13.0 Avg Score : 10.97 | Blue: 0 Yellow: 13 > Elapsed: 1.76sec    \n",
      "Episode: 229 ~ Tstep: 301| Score: 15.0 Avg Score : 11.0 | Blue: 1 Yellow: 16 > Elapsed: 1.78sec     \n",
      "Episode: 230 ~ Tstep: 301| Score: 9.0 Avg Score : 11.01 | Blue: 1 Yellow: 10 > Elapsed: 1.7sec      \n",
      "Episode: 231 ~ Tstep: 301| Score: 17.0 Avg Score : 11.06 | Blue: 0 Yellow: 17 > Elapsed: 1.77sec    \n",
      "Episode: 232 ~ Tstep: 301| Score: 11.0 Avg Score : 11.1 | Blue: 0 Yellow: 11 > Elapsed: 1.74sec     \n",
      "Episode: 233 ~ Tstep: 301| Score: 12.0 Avg Score : 11.17 | Blue: 0 Yellow: 12 > Elapsed: 1.79sec    \n",
      "Episode: 234 ~ Tstep: 301| Score: 11.0 Avg Score : 11.12 | Blue: 0 Yellow: 11 > Elapsed: 1.74sec    \n",
      "Episode: 235 ~ Tstep: 301| Score: 15.0 Avg Score : 11.14 | Blue: 0 Yellow: 15 > Elapsed: 1.74sec    \n",
      "Episode: 236 ~ Tstep: 301| Score: 14.0 Avg Score : 11.21 | Blue: 0 Yellow: 14 > Elapsed: 1.76sec    \n",
      "Episode: 237 ~ Tstep: 301| Score: 5.0 Avg Score : 11.17 | Blue: 0 Yellow: 5 > Elapsed: 1.73sec      \n",
      "Episode: 238 ~ Tstep: 301| Score: 7.0 Avg Score : 11.2 | Blue: 0 Yellow: 7 > Elapsed: 1.76sec       \n",
      "Episode: 239 ~ Tstep: 301| Score: 9.0 Avg Score : 11.11 | Blue: 1 Yellow: 10 > Elapsed: 1.72sec     \n",
      "Episode: 240 ~ Tstep: 301| Score: 14.0 Avg Score : 11.16 | Blue: 0 Yellow: 14 > Elapsed: 1.75sec    \n",
      "Episode: 241 ~ Tstep: 301| Score: 15.0 Avg Score : 11.21 | Blue: 0 Yellow: 15 > Elapsed: 1.78sec    \n",
      "Episode: 242 ~ Tstep: 301| Score: 11.0 Avg Score : 11.29 | Blue: 2 Yellow: 13 > Elapsed: 1.74sec    \n",
      "Episode: 243 ~ Tstep: 301| Score: 13.0 Avg Score : 11.31 | Blue: 0 Yellow: 13 > Elapsed: 1.73sec    \n",
      "Episode: 244 ~ Tstep: 301| Score: 10.0 Avg Score : 11.3 | Blue: 0 Yellow: 10 > Elapsed: 1.76sec     \n",
      "Episode: 245 ~ Tstep: 301| Score: 17.0 Avg Score : 11.42 | Blue: 0 Yellow: 17 > Elapsed: 1.75sec    \n",
      "Episode: 246 ~ Tstep: 301| Score: 16.0 Avg Score : 11.51 | Blue: 0 Yellow: 16 > Elapsed: 1.72sec    \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Episode: 247 ~ Tstep: 301| Score: 13.0 Avg Score : 11.55 | Blue: 0 Yellow: 12 > Elapsed: 1.75sec    \n",
      "Episode: 248 ~ Tstep: 301| Score: 15.0 Avg Score : 11.59 | Blue: 0 Yellow: 15 > Elapsed: 1.75sec    \n",
      "Episode: 249 ~ Tstep: 301| Score: 8.0 Avg Score : 11.59 | Blue: 0 Yellow: 8 > Elapsed: 1.75sec      \n",
      "Episode: 250 ~ Tstep: 301| Score: 16.0 Avg Score : 11.7 | Blue: 0 Yellow: 16 > Elapsed: 1.74sec     \n",
      "Episode: 251 ~ Tstep: 301| Score: 14.0 Avg Score : 11.7 | Blue: 0 Yellow: 14 > Elapsed: 1.72sec     \n",
      "Episode: 252 ~ Tstep: 301| Score: 13.0 Avg Score : 11.78 | Blue: 1 Yellow: 14 > Elapsed: 1.78sec    \n",
      "Episode: 253 ~ Tstep: 301| Score: 12.0 Avg Score : 11.84 | Blue: 2 Yellow: 14 > Elapsed: 1.79sec    \n",
      "Episode: 254 ~ Tstep: 301| Score: 17.0 Avg Score : 11.91 | Blue: 0 Yellow: 17 > Elapsed: 1.7sec     \n",
      "Episode: 255 ~ Tstep: 301| Score: 6.0 Avg Score : 11.82 | Blue: 0 Yellow: 6 > Elapsed: 1.75sec      \n",
      "Episode: 256 ~ Tstep: 301| Score: 13.0 Avg Score : 11.84 | Blue: 0 Yellow: 13 > Elapsed: 1.79sec    \n",
      "Episode: 257 ~ Tstep: 301| Score: 15.0 Avg Score : 11.92 | Blue: 1 Yellow: 16 > Elapsed: 1.75sec    \n",
      "Episode: 258 ~ Tstep: 301| Score: 9.0 Avg Score : 11.9 | Blue: 0 Yellow: 9 > Elapsed: 1.8sec        \n",
      "Episode: 259 ~ Tstep: 301| Score: 5.0 Avg Score : 11.87 | Blue: 1 Yellow: 6 > Elapsed: 1.72sec      \n",
      "Episode: 260 ~ Tstep: 301| Score: 19.0 Avg Score : 11.98 | Blue: 1 Yellow: 20 > Elapsed: 1.76sec    \n",
      "Episode: 261 ~ Tstep: 301| Score: 11.0 Avg Score : 11.99 | Blue: 0 Yellow: 11 > Elapsed: 1.73sec    \n",
      "Episode: 262 ~ Tstep: 301| Score: 13.0 Avg Score : 12.1 | Blue: 0 Yellow: 13 > Elapsed: 1.76sec     \n",
      "Episode: 263 ~ Tstep: 301| Score: 10.0 Avg Score : 12.05 | Blue: 0 Yellow: 10 > Elapsed: 1.73sec    \n",
      "Episode: 264 ~ Tstep: 301| Score: 11.0 Avg Score : 12.03 | Blue: 0 Yellow: 11 > Elapsed: 1.74sec    \n",
      "Episode: 265 ~ Tstep: 301| Score: 14.0 Avg Score : 12.17 | Blue: 0 Yellow: 14 > Elapsed: 1.77sec    \n",
      "Episode: 266 ~ Tstep: 301| Score: 15.0 Avg Score : 12.22 | Blue: 1 Yellow: 16 > Elapsed: 1.74sec    \n",
      "Episode: 267 ~ Tstep: 301| Score: 16.0 Avg Score : 12.24 | Blue: 0 Yellow: 16 > Elapsed: 1.73sec    \n",
      "Episode: 268 ~ Tstep: 301| Score: 10.0 Avg Score : 12.2 | Blue: 0 Yellow: 10 > Elapsed: 1.76sec     \n",
      "Episode: 269 ~ Tstep: 301| Score: 11.0 Avg Score : 12.23 | Blue: 0 Yellow: 11 > Elapsed: 1.75sec    \n",
      "Episode: 270 ~ Tstep: 301| Score: 11.0 Avg Score : 12.24 | Blue: 0 Yellow: 11 > Elapsed: 1.72sec    \n",
      "Episode: 271 ~ Tstep: 301| Score: 6.0 Avg Score : 12.22 | Blue: 0 Yellow: 6 > Elapsed: 1.78sec      \n",
      "Episode: 272 ~ Tstep: 301| Score: 12.0 Avg Score : 12.31 | Blue: 1 Yellow: 13 > Elapsed: 1.74sec    \n",
      "Episode: 273 ~ Tstep: 301| Score: 14.0 Avg Score : 12.36 | Blue: 0 Yellow: 14 > Elapsed: 1.79sec    \n",
      "Episode: 274 ~ Tstep: 301| Score: 10.0 Avg Score : 12.34 | Blue: 1 Yellow: 11 > Elapsed: 1.76sec    \n",
      "Episode: 275 ~ Tstep: 301| Score: 20.0 Avg Score : 12.33 | Blue: 0 Yellow: 20 > Elapsed: 1.76sec    \n",
      "Episode: 276 ~ Tstep: 301| Score: 12.0 Avg Score : 12.38 | Blue: 0 Yellow: 12 > Elapsed: 1.75sec    \n",
      "Episode: 277 ~ Tstep: 301| Score: 14.0 Avg Score : 12.46 | Blue: 2 Yellow: 16 > Elapsed: 1.69sec    \n",
      "Episode: 278 ~ Tstep: 301| Score: 10.0 Avg Score : 12.43 | Blue: 0 Yellow: 10 > Elapsed: 1.72sec    \n",
      "Episode: 279 ~ Tstep: 301| Score: 15.0 Avg Score : 12.47 | Blue: 0 Yellow: 15 > Elapsed: 1.84sec    \n",
      "Episode: 280 ~ Tstep: 301| Score: 14.0 Avg Score : 12.46 | Blue: 0 Yellow: 13 > Elapsed: 1.84sec    \n",
      "Episode: 281 ~ Tstep: 301| Score: 16.0 Avg Score : 12.56 | Blue: 0 Yellow: 16 > Elapsed: 1.8sec     \n",
      "Episode: 282 ~ Tstep: 301| Score: 9.0 Avg Score : 12.51 | Blue: 0 Yellow: 9 > Elapsed: 1.79sec      \n",
      "Episode: 283 ~ Tstep: 301| Score: 6.0 Avg Score : 12.47 | Blue: 0 Yellow: 6 > Elapsed: 1.81sec      \n",
      "Episode: 284 ~ Tstep: 301| Score: 16.0 Avg Score : 12.49 | Blue: 0 Yellow: 16 > Elapsed: 1.95sec    \n",
      "Episode: 285 ~ Tstep: 301| Score: 11.0 Avg Score : 12.56 | Blue: 0 Yellow: 11 > Elapsed: 1.72sec    \n",
      "Episode: 286 ~ Tstep: 301| Score: 1.0 Avg Score : 12.4 | Blue: 0 Yellow: 1 > Elapsed: 1.76sec       \n",
      "Episode: 287 ~ Tstep: 301| Score: 13.0 Avg Score : 12.47 | Blue: 0 Yellow: 13 > Elapsed: 1.77sec    \n",
      "Episode: 288 ~ Tstep: 301| Score: 15.0 Avg Score : 12.58 | Blue: 0 Yellow: 15 > Elapsed: 1.78sec    \n",
      "Episode: 289 ~ Tstep: 301| Score: 13.0 Avg Score : 12.56 | Blue: 2 Yellow: 15 > Elapsed: 1.82sec    \n",
      "Episode: 290 ~ Tstep: 301| Score: 10.0 Avg Score : 12.58 | Blue: 1 Yellow: 11 > Elapsed: 1.79sec    \n",
      "Episode: 291 ~ Tstep: 301| Score: 6.0 Avg Score : 12.52 | Blue: 0 Yellow: 6 > Elapsed: 1.77sec      \n",
      "Episode: 292 ~ Tstep: 301| Score: 8.0 Avg Score : 12.49 | Blue: 0 Yellow: 8 > Elapsed: 1.75sec      \n",
      "Episode: 293 ~ Tstep: 301| Score: 10.0 Avg Score : 12.49 | Blue: 0 Yellow: 10 > Elapsed: 1.73sec    \n",
      "Episode: 294 ~ Tstep: 301| Score: 13.0 Avg Score : 12.44 | Blue: 1 Yellow: 14 > Elapsed: 1.71sec    \n",
      "Episode: 295 ~ Tstep: 301| Score: 13.0 Avg Score : 12.46 | Blue: 0 Yellow: 13 > Elapsed: 1.76sec    \n",
      "Episode: 296 ~ Tstep: 301| Score: 13.0 Avg Score : 12.48 | Blue: 0 Yellow: 13 > Elapsed: 1.78sec    \n",
      "Episode: 297 ~ Tstep: 301| Score: 11.0 Avg Score : 12.43 | Blue: 1 Yellow: 12 > Elapsed: 1.76sec    \n",
      "Episode: 298 ~ Tstep: 301| Score: 7.0 Avg Score : 12.41 | Blue: 1 Yellow: 8 > Elapsed: 1.95sec      \n",
      "Episode: 299 ~ Tstep: 301| Score: 13.0 Avg Score : 12.41 | Blue: 0 Yellow: 13 > Elapsed: 2.52sec    \n",
      "Episode: 300 ~ Tstep: 301| Score: 19.0 Avg Score : 12.46 | Blue: 0 Yellow: 19 > Elapsed: 1.95sec    \n",
      "Episode: 301 ~ Tstep: 301| Score: 2.0 Avg Score : 12.35 | Blue: 0 Yellow: 2 > Elapsed: 1.84sec      \n",
      "Episode: 302 ~ Tstep: 301| Score: 5.0 Avg Score : 12.25 | Blue: 1 Yellow: 6 > Elapsed: 1.92sec      \n",
      "Episode: 303 ~ Tstep: 301| Score: 17.0 Avg Score : 12.3 | Blue: 2 Yellow: 19 > Elapsed: 2.03sec     \n",
      "Episode: 304 ~ Tstep: 301| Score: 9.0 Avg Score : 12.25 | Blue: 1 Yellow: 10 > Elapsed: 1.82sec     \n",
      "Episode: 305 ~ Tstep: 301| Score: 13.0 Avg Score : 12.33 | Blue: 0 Yellow: 13 > Elapsed: 1.88sec    \n",
      "Episode: 306 ~ Tstep: 301| Score: 6.0 Avg Score : 12.23 | Blue: 0 Yellow: 6 > Elapsed: 1.85sec      \n",
      "Episode: 307 ~ Tstep: 301| Score: 6.0 Avg Score : 12.12 | Blue: 0 Yellow: 6 > Elapsed: 1.84sec      \n",
      "Episode: 308 ~ Tstep: 301| Score: 11.0 Avg Score : 12.1 | Blue: 0 Yellow: 11 > Elapsed: 1.95sec     \n",
      "Episode: 309 ~ Tstep: 301| Score: 5.0 Avg Score : 12.02 | Blue: 0 Yellow: 5 > Elapsed: 1.8sec       \n",
      "Episode: 310 ~ Tstep: 301| Score: 14.0 Avg Score : 12.01 | Blue: 0 Yellow: 14 > Elapsed: 1.86sec    \n",
      "Episode: 311 ~ Tstep: 301| Score: 6.0 Avg Score : 11.92 | Blue: 0 Yellow: 6 > Elapsed: 1.92sec      \n",
      "Episode: 312 ~ Tstep: 301| Score: 17.0 Avg Score : 11.94 | Blue: 0 Yellow: 17 > Elapsed: 1.9sec     \n",
      "Episode: 313 ~ Tstep: 301| Score: 20.0 Avg Score : 12.0 | Blue: 0 Yellow: 20 > Elapsed: 2.36sec     \n",
      "Episode: 314 ~ Tstep: 301| Score: 12.0 Avg Score : 11.96 | Blue: 0 Yellow: 12 > Elapsed: 1.91sec    \n",
      "Episode: 315 ~ Tstep: 301| Score: 12.0 Avg Score : 11.89 | Blue: 0 Yellow: 12 > Elapsed: 1.94sec    \n",
      "Episode: 316 ~ Tstep: 301| Score: 8.0 Avg Score : 11.89 | Blue: 1 Yellow: 9 > Elapsed: 1.81sec      \n",
      "Episode: 317 ~ Tstep: 301| Score: 20.0 Avg Score : 11.99 | Blue: 1 Yellow: 21 > Elapsed: 1.81sec    \n",
      "Episode: 318 ~ Tstep: 301| Score: 16.0 Avg Score : 11.99 | Blue: 1 Yellow: 17 > Elapsed: 1.92sec    \n",
      "Episode: 319 ~ Tstep: 301| Score: 23.0 Avg Score : 12.08 | Blue: 0 Yellow: 23 > Elapsed: 1.85sec    \n",
      "Episode: 320 ~ Tstep: 301| Score: 11.0 Avg Score : 12.06 | Blue: 0 Yellow: 11 > Elapsed: 1.83sec    \n",
      "Episode: 321 ~ Tstep: 301| Score: 22.0 Avg Score : 12.16 | Blue: 0 Yellow: 22 > Elapsed: 1.83sec    \n",
      "Episode: 322 ~ Tstep: 301| Score: 19.0 Avg Score : 12.2 | Blue: 0 Yellow: 19 > Elapsed: 1.81sec     \n",
      "Episode: 323 ~ Tstep: 301| Score: 8.0 Avg Score : 12.13 | Blue: 1 Yellow: 9 > Elapsed: 1.81sec      \n",
      "Episode: 324 ~ Tstep: 301| Score: 12.0 Avg Score : 12.1 | Blue: 0 Yellow: 12 > Elapsed: 1.85sec     \n",
      "Episode: 325 ~ Tstep: 301| Score: 12.0 Avg Score : 12.13 | Blue: 1 Yellow: 13 > Elapsed: 1.81sec    \n",
      "Episode: 326 ~ Tstep: 301| Score: 16.0 Avg Score : 12.14 | Blue: 0 Yellow: 16 > Elapsed: 1.84sec    \n",
      "Episode: 327 ~ Tstep: 301| Score: 11.0 Avg Score : 12.14 | Blue: 0 Yellow: 11 > Elapsed: 1.88sec    \n",
      "Episode: 328 ~ Tstep: 301| Score: 9.0 Avg Score : 12.1 | Blue: 0 Yellow: 9 > Elapsed: 1.87sec       \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Episode: 329 ~ Tstep: 301| Score: 8.0 Avg Score : 12.03 | Blue: 0 Yellow: 8 > Elapsed: 1.85sec      \n",
      "Episode: 330 ~ Tstep: 301| Score: 21.0 Avg Score : 12.15 | Blue: 0 Yellow: 21 > Elapsed: 1.85sec    \n",
      "Episode: 331 ~ Tstep: 301| Score: 8.0 Avg Score : 12.06 | Blue: 0 Yellow: 8 > Elapsed: 1.89sec      \n",
      "Episode: 332 ~ Tstep: 301| Score: 13.0 Avg Score : 12.08 | Blue: 1 Yellow: 14 > Elapsed: 1.88sec    \n",
      "Episode: 333 ~ Tstep: 301| Score: 13.0 Avg Score : 12.09 | Blue: 0 Yellow: 13 > Elapsed: 1.82sec    \n",
      "Episode: 334 ~ Tstep: 301| Score: 12.0 Avg Score : 12.1 | Blue: 2 Yellow: 14 > Elapsed: 1.81sec     \n",
      "Episode: 335 ~ Tstep: 301| Score: 18.0 Avg Score : 12.13 | Blue: 0 Yellow: 18 > Elapsed: 1.89sec    \n",
      "Episode: 336 ~ Tstep: 301| Score: 17.0 Avg Score : 12.16 | Blue: 0 Yellow: 17 > Elapsed: 1.87sec    \n",
      "Episode: 337 ~ Tstep: 301| Score: 16.0 Avg Score : 12.27 | Blue: 0 Yellow: 16 > Elapsed: 1.87sec    \n",
      "Episode: 338 ~ Tstep: 301| Score: 10.0 Avg Score : 12.3 | Blue: 0 Yellow: 10 > Elapsed: 1.79sec     \n",
      "Episode: 339 ~ Tstep: 301| Score: 17.0 Avg Score : 12.38 | Blue: 0 Yellow: 17 > Elapsed: 1.79sec    \n",
      "Episode: 340 ~ Tstep: 301| Score: 3.0 Avg Score : 12.27 | Blue: 0 Yellow: 3 > Elapsed: 1.84sec      \n",
      "Episode: 341 ~ Tstep: 301| Score: 11.0 Avg Score : 12.23 | Blue: 1 Yellow: 12 > Elapsed: 1.95sec    \n",
      "Episode: 342 ~ Tstep: 301| Score: 9.0 Avg Score : 12.21 | Blue: 1 Yellow: 10 > Elapsed: 1.76sec     \n",
      "Episode: 343 ~ Tstep: 301| Score: 14.0 Avg Score : 12.22 | Blue: 0 Yellow: 14 > Elapsed: 1.91sec    \n",
      "Episode: 344 ~ Tstep: 301| Score: 7.0 Avg Score : 12.19 | Blue: 1 Yellow: 8 > Elapsed: 1.77sec      \n",
      "Episode: 345 ~ Tstep: 301| Score: 10.0 Avg Score : 12.12 | Blue: 0 Yellow: 10 > Elapsed: 1.79sec    \n",
      "Episode: 346 ~ Tstep: 301| Score: 8.0 Avg Score : 12.04 | Blue: 0 Yellow: 8 > Elapsed: 1.97sec      \n",
      "Episode: 347 ~ Tstep: 301| Score: 14.0 Avg Score : 12.05 | Blue: 0 Yellow: 14 > Elapsed: 1.83sec    \n",
      "Episode: 348 ~ Tstep: 301| Score: 11.0 Avg Score : 12.01 | Blue: 0 Yellow: 11 > Elapsed: 1.76sec    \n",
      "Episode: 349 ~ Tstep: 301| Score: 12.0 Avg Score : 12.05 | Blue: 0 Yellow: 12 > Elapsed: 1.83sec    \n",
      "Episode: 350 ~ Tstep: 301| Score: 12.0 Avg Score : 12.01 | Blue: 0 Yellow: 12 > Elapsed: 1.93sec    \n",
      "Episode: 351 ~ Tstep: 301| Score: 11.0 Avg Score : 11.98 | Blue: 2 Yellow: 13 > Elapsed: 2.06sec    \n",
      "Episode: 352 ~ Tstep: 301| Score: 11.0 Avg Score : 11.96 | Blue: 1 Yellow: 12 > Elapsed: 2.05sec    \n",
      "Episode: 353 ~ Tstep: 301| Score: 11.0 Avg Score : 11.95 | Blue: 0 Yellow: 11 > Elapsed: 1.97sec    \n",
      "Episode: 354 ~ Tstep: 301| Score: 21.0 Avg Score : 11.99 | Blue: 0 Yellow: 21 > Elapsed: 1.82sec    \n",
      "Episode: 355 ~ Tstep: 301| Score: 10.0 Avg Score : 12.03 | Blue: 1 Yellow: 11 > Elapsed: 1.83sec    \n",
      "Episode: 356 ~ Tstep: 301| Score: 12.0 Avg Score : 12.02 | Blue: 0 Yellow: 12 > Elapsed: 1.89sec    \n",
      "Episode: 357 ~ Tstep: 301| Score: 18.0 Avg Score : 12.05 | Blue: 0 Yellow: 18 > Elapsed: 1.87sec    \n",
      "Episode: 358 ~ Tstep: 301| Score: 16.0 Avg Score : 12.12 | Blue: 0 Yellow: 16 > Elapsed: 1.9sec     \n",
      "Episode: 359 ~ Tstep: 301| Score: 18.0 Avg Score : 12.25 | Blue: 0 Yellow: 18 > Elapsed: 1.83sec    \n",
      "Episode: 360 ~ Tstep: 301| Score: 18.0 Avg Score : 12.24 | Blue: 0 Yellow: 18 > Elapsed: 1.87sec    \n",
      "Episode: 361 ~ Tstep: 301| Score: 17.0 Avg Score : 12.3 | Blue: 0 Yellow: 17 > Elapsed: 1.77sec     \n",
      "Episode: 362 ~ Tstep: 301| Score: 8.0 Avg Score : 12.25 | Blue: 0 Yellow: 8 > Elapsed: 1.89sec      \n",
      "Episode: 363 ~ Tstep: 301| Score: 21.0 Avg Score : 12.36 | Blue: 0 Yellow: 21 > Elapsed: 1.82sec    \n",
      "Episode: 364 ~ Tstep: 301| Score: 14.0 Avg Score : 12.39 | Blue: 0 Yellow: 14 > Elapsed: 1.92sec    \n",
      "Episode: 365 ~ Tstep: 301| Score: 21.0 Avg Score : 12.46 | Blue: 1 Yellow: 22 > Elapsed: 1.76sec    \n",
      "Episode: 366 ~ Tstep: 301| Score: 11.0 Avg Score : 12.42 | Blue: 0 Yellow: 11 > Elapsed: 1.83sec    \n",
      "Episode: 367 ~ Tstep: 301| Score: 16.0 Avg Score : 12.42 | Blue: 0 Yellow: 16 > Elapsed: 1.95sec    \n",
      "Episode: 368 ~ Tstep: 301| Score: 15.0 Avg Score : 12.47 | Blue: 0 Yellow: 15 > Elapsed: 1.87sec    \n",
      "Episode: 369 ~ Tstep: 301| Score: 18.0 Avg Score : 12.54 | Blue: 0 Yellow: 18 > Elapsed: 1.89sec    \n",
      "Episode: 370 ~ Tstep: 301| Score: 12.0 Avg Score : 12.55 | Blue: 0 Yellow: 12 > Elapsed: 1.83sec    \n",
      "Episode: 371 ~ Tstep: 301| Score: 13.0 Avg Score : 12.62 | Blue: 0 Yellow: 13 > Elapsed: 1.92sec    \n",
      "Episode: 372 ~ Tstep: 301| Score: 13.0 Avg Score : 12.63 | Blue: 1 Yellow: 14 > Elapsed: 1.92sec    \n",
      "Episode: 373 ~ Tstep: 301| Score: 12.0 Avg Score : 12.61 | Blue: 1 Yellow: 13 > Elapsed: 1.88sec    \n",
      "Episode: 374 ~ Tstep: 301| Score: 16.0 Avg Score : 12.67 | Blue: 0 Yellow: 16 > Elapsed: 1.92sec    \n",
      "Episode: 375 ~ Tstep: 301| Score: 18.0 Avg Score : 12.65 | Blue: 0 Yellow: 18 > Elapsed: 1.95sec    \n",
      "Episode: 376 ~ Tstep: 301| Score: 19.0 Avg Score : 12.72 | Blue: 0 Yellow: 19 > Elapsed: 1.96sec    \n",
      "Episode: 377 ~ Tstep: 301| Score: 10.0 Avg Score : 12.68 | Blue: 1 Yellow: 11 > Elapsed: 2.02sec    \n",
      "Episode: 378 ~ Tstep: 301| Score: 21.0 Avg Score : 12.79 | Blue: 0 Yellow: 21 > Elapsed: 1.91sec    \n",
      "Episode: 379 ~ Tstep: 301| Score: 17.0 Avg Score : 12.81 | Blue: 1 Yellow: 18 > Elapsed: 1.94sec    \n",
      "Episode: 380 ~ Tstep: 301| Score: 19.0 Avg Score : 12.86 | Blue: 0 Yellow: 19 > Elapsed: 1.97sec    \n",
      "Episode: 381 ~ Tstep: 301| Score: 13.0 Avg Score : 12.83 | Blue: 0 Yellow: 13 > Elapsed: 2.04sec    \n",
      "Episode: 382 ~ Tstep: 301| Score: 22.0 Avg Score : 12.96 | Blue: 0 Yellow: 22 > Elapsed: 2.34sec    \n",
      "Episode: 383 ~ Tstep: 301| Score: 18.0 Avg Score : 13.08 | Blue: 1 Yellow: 19 > Elapsed: 1.97sec    \n",
      "Score Threshold of 13.0 has been achieved - Saving a snapshot of the Mdoel ....\n",
      "Episode: 384 ~ Tstep: 301| Score: 16.0 Avg Score : 13.08 | Blue: 1 Yellow: 17 > Elapsed: 1.95sec    \n",
      "Score Threshold of 13.0 has been achieved - Saving a snapshot of the Mdoel ....\n",
      "Episode: 385 ~ Tstep: 301| Score: 22.0 Avg Score : 13.19 | Blue: 0 Yellow: 22 > Elapsed: 2.0sec     \n",
      "Episode: 386 ~ Tstep: 301| Score: 18.0 Avg Score : 13.36 | Blue: 0 Yellow: 18 > Elapsed: 1.98sec    \n",
      "Episode: 387 ~ Tstep: 301| Score: 11.0 Avg Score : 13.34 | Blue: 0 Yellow: 11 > Elapsed: 2.01sec    \n",
      "Episode: 388 ~ Tstep: 301| Score: 1.0 Avg Score : 13.2 | Blue: 0 Yellow: 1 > Elapsed: 1.99sec       \n",
      "Episode: 389 ~ Tstep: 301| Score: 19.0 Avg Score : 13.26 | Blue: 0 Yellow: 19 > Elapsed: 2.02sec    \n",
      "Episode: 390 ~ Tstep: 301| Score: 12.0 Avg Score : 13.28 | Blue: 0 Yellow: 12 > Elapsed: 1.97sec    \n",
      "Episode: 391 ~ Tstep: 301| Score: 11.0 Avg Score : 13.33 | Blue: 0 Yellow: 11 > Elapsed: 2.04sec    \n",
      "Episode: 392 ~ Tstep: 301| Score: 14.0 Avg Score : 13.39 | Blue: 0 Yellow: 14 > Elapsed: 2.05sec    \n",
      "Episode: 393 ~ Tstep: 301| Score: 17.0 Avg Score : 13.46 | Blue: 0 Yellow: 17 > Elapsed: 1.97sec    \n",
      "Episode: 394 ~ Tstep: 301| Score: 6.0 Avg Score : 13.39 | Blue: 1 Yellow: 7 > Elapsed: 1.94sec      \n",
      "Episode: 395 ~ Tstep: 301| Score: 12.0 Avg Score : 13.38 | Blue: 1 Yellow: 13 > Elapsed: 1.99sec    \n",
      "Episode: 396 ~ Tstep: 301| Score: 12.0 Avg Score : 13.37 | Blue: 1 Yellow: 13 > Elapsed: 2.04sec    \n",
      "Episode: 397 ~ Tstep: 301| Score: 9.0 Avg Score : 13.35 | Blue: 0 Yellow: 9 > Elapsed: 2.05sec      \n",
      "Episode: 398 ~ Tstep: 301| Score: 16.0 Avg Score : 13.44 | Blue: 1 Yellow: 17 > Elapsed: 1.86sec    \n",
      "Episode: 399 ~ Tstep: 301| Score: 12.0 Avg Score : 13.43 | Blue: 0 Yellow: 12 > Elapsed: 1.82sec    \n",
      "Episode: 400 ~ Tstep: 301| Score: 12.0 Avg Score : 13.36 | Blue: 0 Yellow: 12 > Elapsed: 1.78sec    \n",
      "Episode: 401 ~ Tstep: 301| Score: 18.0 Avg Score : 13.52 | Blue: 0 Yellow: 18 > Elapsed: 1.81sec    \n",
      "Episode: 402 ~ Tstep: 301| Score: 3.0 Avg Score : 13.5 | Blue: 1 Yellow: 4 > Elapsed: 1.75sec       \n",
      "Episode: 403 ~ Tstep: 301| Score: 13.0 Avg Score : 13.46 | Blue: 0 Yellow: 13 > Elapsed: 1.79sec    \n",
      "Episode: 404 ~ Tstep: 301| Score: 17.0 Avg Score : 13.54 | Blue: 1 Yellow: 18 > Elapsed: 1.79sec    \n",
      "Episode: 405 ~ Tstep: 301| Score: 16.0 Avg Score : 13.57 | Blue: 0 Yellow: 16 > Elapsed: 1.81sec    \n",
      "Episode: 406 ~ Tstep: 301| Score: 18.0 Avg Score : 13.69 | Blue: 0 Yellow: 18 > Elapsed: 1.79sec    \n",
      "Episode: 407 ~ Tstep: 301| Score: 4.0 Avg Score : 13.67 | Blue: 2 Yellow: 6 > Elapsed: 1.77sec      \n",
      "Episode: 408 ~ Tstep: 301| Score: 17.0 Avg Score : 13.73 | Blue: 0 Yellow: 17 > Elapsed: 1.81sec    \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Episode: 409 ~ Tstep: 301| Score: 15.0 Avg Score : 13.83 | Blue: 0 Yellow: 15 > Elapsed: 1.75sec    \n",
      "Episode: 410 ~ Tstep: 301| Score: 18.0 Avg Score : 13.87 | Blue: 0 Yellow: 18 > Elapsed: 1.78sec    \n",
      "Episode: 411 ~ Tstep: 301| Score: 11.0 Avg Score : 13.92 | Blue: 0 Yellow: 11 > Elapsed: 1.81sec    \n",
      "Episode: 412 ~ Tstep: 301| Score: 12.0 Avg Score : 13.87 | Blue: 0 Yellow: 12 > Elapsed: 1.84sec    \n",
      "Episode: 413 ~ Tstep: 301| Score: 15.0 Avg Score : 13.82 | Blue: 0 Yellow: 15 > Elapsed: 1.8sec     \n",
      "Episode: 414 ~ Tstep: 301| Score: 3.0 Avg Score : 13.73 | Blue: 1 Yellow: 4 > Elapsed: 1.77sec      \n",
      "Episode: 415 ~ Tstep: 301| Score: 14.0 Avg Score : 13.75 | Blue: 2 Yellow: 16 > Elapsed: 1.87sec    \n",
      "Episode: 416 ~ Tstep: 301| Score: 18.0 Avg Score : 13.85 | Blue: 0 Yellow: 18 > Elapsed: 1.89sec    \n",
      "Episode: 417 ~ Tstep: 301| Score: 14.0 Avg Score : 13.79 | Blue: 0 Yellow: 14 > Elapsed: 1.79sec    \n",
      "Episode: 418 ~ Tstep: 301| Score: 10.0 Avg Score : 13.73 | Blue: 0 Yellow: 10 > Elapsed: 1.93sec    \n",
      "Episode: 419 ~ Tstep: 301| Score: 15.0 Avg Score : 13.65 | Blue: 0 Yellow: 15 > Elapsed: 1.76sec    \n",
      "Episode: 420 ~ Tstep: 301| Score: 8.0 Avg Score : 13.62 | Blue: 1 Yellow: 9 > Elapsed: 1.72sec      \n",
      "Episode: 421 ~ Tstep: 301| Score: 11.0 Avg Score : 13.51 | Blue: 2 Yellow: 13 > Elapsed: 1.74sec    \n",
      "Episode: 422 ~ Tstep: 301| Score: 9.0 Avg Score : 13.41 | Blue: 0 Yellow: 9 > Elapsed: 1.81sec      \n",
      "Episode: 423 ~ Tstep: 301| Score: 15.0 Avg Score : 13.48 | Blue: 0 Yellow: 14 > Elapsed: 1.81sec    \n",
      "Episode: 424 ~ Tstep: 301| Score: 15.0 Avg Score : 13.51 | Blue: 1 Yellow: 16 > Elapsed: 1.81sec    \n",
      "Episode: 425 ~ Tstep: 301| Score: 2.0 Avg Score : 13.41 | Blue: 0 Yellow: 2 > Elapsed: 1.78sec      \n",
      "Episode: 426 ~ Tstep: 301| Score: 16.0 Avg Score : 13.41 | Blue: 0 Yellow: 16 > Elapsed: 1.84sec    \n",
      "Episode: 427 ~ Tstep: 301| Score: 14.0 Avg Score : 13.44 | Blue: 0 Yellow: 14 > Elapsed: 1.73sec    \n",
      "Episode: 428 ~ Tstep: 301| Score: 15.0 Avg Score : 13.5 | Blue: 0 Yellow: 15 > Elapsed: 1.81sec     \n",
      "Episode: 429 ~ Tstep: 301| Score: 18.0 Avg Score : 13.6 | Blue: 0 Yellow: 18 > Elapsed: 1.75sec     \n",
      "Episode: 430 ~ Tstep: 301| Score: 15.0 Avg Score : 13.54 | Blue: 1 Yellow: 16 > Elapsed: 1.73sec    \n",
      "Episode: 431 ~ Tstep: 301| Score: 19.0 Avg Score : 13.65 | Blue: 0 Yellow: 19 > Elapsed: 1.73sec    \n",
      "Episode: 432 ~ Tstep: 301| Score: 12.0 Avg Score : 13.64 | Blue: 1 Yellow: 13 > Elapsed: 1.8sec     \n",
      "Episode: 433 ~ Tstep: 301| Score: 12.0 Avg Score : 13.63 | Blue: 0 Yellow: 12 > Elapsed: 1.76sec    \n",
      "Episode: 434 ~ Tstep: 301| Score: 5.0 Avg Score : 13.56 | Blue: 1 Yellow: 6 > Elapsed: 1.84sec      \n",
      "Episode: 435 ~ Tstep: 301| Score: 23.0 Avg Score : 13.61 | Blue: 2 Yellow: 25 > Elapsed: 1.78sec    \n",
      "Episode: 436 ~ Tstep: 301| Score: 12.0 Avg Score : 13.56 | Blue: 0 Yellow: 12 > Elapsed: 1.76sec    \n",
      "Episode: 437 ~ Tstep: 301| Score: 15.0 Avg Score : 13.55 | Blue: 0 Yellow: 15 > Elapsed: 1.76sec    \n",
      "Episode: 438 ~ Tstep: 301| Score: 16.0 Avg Score : 13.61 | Blue: 1 Yellow: 17 > Elapsed: 1.82sec    \n",
      "Episode: 439 ~ Tstep: 301| Score: 16.0 Avg Score : 13.6 | Blue: 0 Yellow: 16 > Elapsed: 1.77sec     \n",
      "Episode: 440 ~ Tstep: 301| Score: 10.0 Avg Score : 13.67 | Blue: 0 Yellow: 10 > Elapsed: 1.75sec    \n",
      "Episode: 441 ~ Tstep: 301| Score: 8.0 Avg Score : 13.64 | Blue: 1 Yellow: 9 > Elapsed: 1.76sec      \n",
      "Episode: 442 ~ Tstep: 301| Score: 14.0 Avg Score : 13.69 | Blue: 1 Yellow: 15 > Elapsed: 1.8sec     \n",
      "Episode: 443 ~ Tstep: 301| Score: 15.0 Avg Score : 13.7 | Blue: 0 Yellow: 15 > Elapsed: 1.78sec     \n",
      "Episode: 444 ~ Tstep: 301| Score: 16.0 Avg Score : 13.79 | Blue: 0 Yellow: 16 > Elapsed: 1.69sec    \n",
      "Episode: 445 ~ Tstep: 301| Score: 23.0 Avg Score : 13.92 | Blue: 1 Yellow: 24 > Elapsed: 1.79sec    \n",
      "Episode: 446 ~ Tstep: 301| Score: 15.0 Avg Score : 13.99 | Blue: 0 Yellow: 15 > Elapsed: 1.78sec    \n",
      "Episode: 447 ~ Tstep: 301| Score: 13.0 Avg Score : 13.98 | Blue: 0 Yellow: 13 > Elapsed: 1.72sec    \n",
      "Episode: 448 ~ Tstep: 301| Score: 11.0 Avg Score : 13.98 | Blue: 0 Yellow: 11 > Elapsed: 1.77sec    \n",
      "Episode: 449 ~ Tstep: 301| Score: 14.0 Avg Score : 14.0 | Blue: 0 Yellow: 14 > Elapsed: 1.77sec     \n",
      "Model has consistently maintained a score much above the threshold of 13.0 over 100 episodes\n",
      "\n",
      "... Stopping training....\n",
      "\n",
      "\n",
      "Total Elapsed Train time: 1374.884 seconds | 22.91 minutes\n"
     ]
    }
   ],
   "source": [
    "train_start_time = time.time()\n",
    "\n",
    "try: \n",
    "    for i in range(num_episodes):\n",
    "        \"\"\"\n",
    "            At the beginning of the Episode ->\n",
    "                1> Initialise the episode_score = 0\n",
    "                2> Reset the Environment State\n",
    "        \"\"\"\n",
    "        score = 0                                         # Initialise Agent's episode score\n",
    "        inspect_score = 0                                 # Initilaise Avg Score over 'inpect_episodes'\n",
    "        env_info = env.reset(train_mode=True)[brain_name] # reset the environment\n",
    "        state = env_info.vector_observations[0]           # get the curent state\n",
    "        \n",
    "        # epsilon = 1.0 at first Episode -> decay episodic epsilon over episodes\n",
    "        epi_epsilon = epi_epsilon*episodic_epsilon_decay  \n",
    "        epi_epsilon = 0.0 if epi_epsilon < epsilon_thresh else epi_epsilon # threshold to stop Agent action Exploration\n",
    "        \n",
    "        epsilon = epi_epsilon # init epsilon for one Episode which wil decay over timesteps in an Episode\n",
    "        \n",
    "        done = False\n",
    "        episode_start_time = time.time()\n",
    "        timestep = 1\n",
    "        yellow, blue = 0, 0\n",
    "\n",
    "        # Keep running the environment simulation until the Terminal State has been reached\n",
    "        # EPISODE RUN\n",
    "        while not done:\n",
    "            \n",
    "            action, q_estimate = None, None\n",
    "            \n",
    "            if IS_PRIORITIZED_REPLAY:\n",
    "                # Agent WITH Prioritized Action Replay\n",
    "                action, q_estimate = agent.act(state, epsilon) # Agent Action\n",
    "            else:\n",
    "                # Agent without any Prioritized Action Replay\n",
    "                action = agent.act(state, epsilon) # action = agent.act(state)\n",
    "            \n",
    "            action = action.astype('int32')\n",
    "            \n",
    "            env_info = env.step(action)[brain_name] # send Agent action to the Environment\n",
    "\n",
    "            # Obtain:> 'next_state', 'reward', 'done' params for the Agent-Environment interaction\n",
    "            next_state = env_info.vector_observations[0]   \n",
    "            reward = env_info.rewards[0]                   \n",
    "            done = env_info.local_done[0]                  \n",
    "\n",
    "            # Agent Step to Update Learning-parameters\n",
    "            \n",
    "            if IS_PRIORITIZED_REPLAY:\n",
    "                # Agent WITH Prioritized Action Replay\n",
    "                agent.step( q_estimate, state, action, reward, next_state, done )\n",
    "\n",
    "            else:\n",
    "                # Agent WITHOUT Prioritized Action Replay\n",
    "                agent.step( state, action, reward, next_state, done )\n",
    "\n",
    "            # check for Yellow Banana or Blue\n",
    "            if reward > 0:\n",
    "                yellow += 1\n",
    "            elif reward < 0:\n",
    "                blue += 1\n",
    "\n",
    "            # Update the timestep-score\n",
    "            score += reward\n",
    "\n",
    "            # State-transition: update current state to next state\n",
    "            state = next_state\n",
    "\n",
    "            # Epsion (Greedy parameter) decay over timesteps & capping to low_thresh of 0.1\n",
    "            epsilon = epsilon * epsilon_decay\n",
    "            epsilon = 0.0001 if epsilon < epsilon_thresh else epsilon  \n",
    "\n",
    "            timestep += 1\n",
    "\n",
    "        # One EPISODE RUN has completed\n",
    "        episode_score.append(score)\n",
    "\n",
    "        # Update the Avg Score in the Past 'inspect_episodes'\n",
    "        if inspect_episodes < i + 1: \n",
    "            inspect_score = np.mean( episode_score[-inspect_episodes:] )\n",
    "            avg_score_per_inspect_episodes.append( inspect_score )\n",
    "\n",
    "            # inspect_log =\"\\rAverage Score over {} episodes : {}\".format(inspect_episodes, inspect_score)\n",
    "            # inspect_log = inspect_log + \" \"*(100 - len(inspect_log))\n",
    "            # print( inspect_log, end = \"\\r\" )\n",
    "\n",
    "        episode_end_time = time.time()\n",
    "\n",
    "        log=\"Episode: {} ~ Tstep: {}| Score: {} Avg Score : {} | Blue: {} Yellow: {} > Elapsed: {}sec\".format(\n",
    "                                                                           i+1, timestep,\n",
    "                                                                           score, inspect_score,\n",
    "                                                                           blue, yellow,\n",
    "                                                                           np.round(episode_end_time-episode_start_time,2)\n",
    "                                                                           )\n",
    "        log = log + \" \"*(100 - len(log))\n",
    "        #print( log, end = \"\\r\" )\n",
    "        print(log)\n",
    "\n",
    "        # Train Checkpoint / Termination Condition\n",
    "        if inspect_score_thresh <= inspect_score and num_checkpoint < 3:\n",
    "            checkpoint(agent, num_checkpoint)\n",
    "            print(\"Score Threshold of {} has been achieved - Saving a snapshot of the Mdoel ....\"\n",
    "                  .format(inspect_score_thresh)\n",
    "                 )\n",
    "            num_checkpoint += 1\n",
    "\n",
    "        elif 14 <= inspect_score and 2 < num_checkpoint:\n",
    "            checkpoint(agent, num_checkpoint+1)\n",
    "            print(\"Model has consistently maintained a score much above the threshold of {} over {} episodes\".format(inspect_score_thresh,\n",
    "                                                                                                                     inspect_episodes\n",
    "                                                                                                                    )\n",
    "                 )\n",
    "            print(\"\\n... Stopping training....\\n\")\n",
    "            break\n",
    "\n",
    "        time.sleep(1.25) # Sleep of ~1 second\n",
    "        \n",
    "except KeyboardInterrupt:\n",
    "    print(\"\\nKeyboardInterrupt exception -> Aborting RL Agent Train\")\n",
    "    \n",
    "train_end_time = time.time()\n",
    "train_time = np.round(train_end_time-train_start_time, 3)\n",
    "print(\"\\nTotal Elapsed Train time: {} seconds | {} minutes\".format( train_time, np.round(train_time/60.0, 2) )\n",
    "     )\n",
    "\n",
    "# Close the Environment\n",
    "env.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(449, 349)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#len(agent.buffer), len(episode_score), len(avg_score_per_inspect_episodes)\n",
    "len(episode_score), len(avg_score_per_inspect_episodes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "plt.figure(figsize=(8, 6))\n",
    "plt.plot( [i for i in range(len(episode_score)) ], episode_score)\n",
    "plt.ylabel(\"Agent Reward - Yellow Bananas Collected\")\n",
    "plt.xlabel(\"Episodes\")\n",
    "plt.title(\"Train Summary - DQN Agent in Navigation / Object Avoidance Env.\")\n",
    "\n",
    "plt.xticks(np.arange(0, len(episode_score)+1, 50))\n",
    "plt.yticks(np.arange(0, max(episode_score)+1, 1.0))\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "plt.figure(figsize=(8, 6))\n",
    "plt.plot( [i for i in range(len(avg_score_per_inspect_episodes)) ], avg_score_per_inspect_episodes)\n",
    "\n",
    "plt.ylabel(\"Average Agent Reward - Yellow Bananas Collected\")\n",
    "plt.xlabel(\"Episodes\")\n",
    "plt.title(\"Avg. Reward Episode Window Train Summary - DQN Agent in Object Avoidance Env.\")\n",
    "\n",
    "plt.xticks(np.arange(0, len(avg_score_per_inspect_episodes)+1, 50))\n",
    "plt.yticks(np.arange(0, max(avg_score_per_inspect_episodes)+1, 1.0))\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "plt.figure(figsize=(8, 6))\n",
    "\n",
    "obs_diff = len(episode_score) - len(avg_score_per_inspect_episodes)\n",
    "\n",
    "plt.plot( [i for i in range(len(episode_score)) ], episode_score)\n",
    "plt.plot( [i for i in range(len(episode_score)) ], obs_diff*[None] + avg_score_per_inspect_episodes,\n",
    "          linewidth=3.0\n",
    "        )\n",
    "\n",
    "plt.ylabel(\"Agent Reward - Yellow Bananas Collected\")\n",
    "plt.xlabel(\"Episodes\")\n",
    "plt.title(\"Train Summary - DQN Agent in Object Avoidance Env.\")\n",
    "\n",
    "plt.legend([\"Reward per Episode\", \"Rolling Avg. Reward over 100 Episodes\"], loc =\"lower right\")\n",
    "\n",
    "plt.xticks(np.arange(0, len(episode_score)+1, 50))\n",
    "plt.yticks(np.arange(0, max(episode_score)+1, 1.0))\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h4>Store the <em>'Episode Score'</em> for the specific trial</h4>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.savez(\"./Episode_Scores/Episode_Scores_{}_{}.npz\".format(model_ver, trial_ver), episode_score)\n",
    "np.savez(\"./Episode_Scores/Inspect_Episode_Scores_{}_{}.npz\".format(model_ver, trial_ver), avg_score_per_inspect_episodes)"
   ]
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}