Google_Dopamine_(LIVE).ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Google Dopamine (LIVE)",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "[View in Colaboratory](https://colab.research.google.com/github/jonbaer/googlecolab/blob/master/Google_Dopamine_(LIVE).ipynb)"
      ]
    },
    {
      "metadata": {
        "id": "mLcmGG387N5t",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "\n",
        "# Motivations\n",
        "\n",
        "![alt text](https://camo.githubusercontent.com/96b445af53319b5f70795e42ca29263c2b4e161e/68747470733a2f2f676f6f676c652e6769746875622e696f2f646f70616d696e652f696d616765732f646f70616d696e655f6c6f676f2e706e67)\n",
        "\n",
        "![alt text](https://www.skynettoday.com/content/editorials/images/is-alphago-zero-overrated/history.png)\n",
        "\n",
        "- Deep Reinforcment learning is responsible for the 2 biggest AI victories most recently (The Deep Q Network, AlphaGo, OpenAI Five)\n",
        "- Deep Q started the surge of interest in Deep RL\n",
        "- Deep Q was so impressive that Google bought DeepMind for half a billion US dollars to leverage it for their products\n",
        "- Deep Q exceled because of 3 (at the time) state-of-the art techniques; replay memory, large-scale distributed training, and distributional modeling methods\n",
        "- The time and attention it took Google researchers to develop these techniques made them realize they needed a framework that fit some key pain points\n",
        "\n",
        "### Wait, Siraj. What's replay memory?\n",
        "- To perform experience replay, an agent will  store it's experiences t=(s,a,r,st+1)\n",
        "- So instead of running Q-learning on state/action pairs as they occur during simulation or actual experience, the system stores the data discovered for [state, action, reward, next_state] - typically in a large table. \n",
        "Note this does not store associated values - this is the raw data to feed into action-value calculations later.\n",
        "- The learning phase is then logically separate from gaining experience, and based on taking random samples from this table. \n",
        "- You still want to interleave the two processes - acting and learning - because improving the policy will lead to different behaviour that should explore actions closer to optimal ones, and you want to learn from those.\n",
        "\n",
        "### Can we learn Deep reinforcement learning now lol?\n",
        "- Neural networks are the agent that learns to map state-action pairs to rewards. \n",
        "- Thats all for now. We need to continue with pure RL before deep RL\n",
        "\n",
        "# Paint Points include\n",
        "\n",
        "- Existing reinforcement learning frameworks are not flexible. It takes time to make changes, and this kind of research requires rapid experimentation\n",
        "- Code reproducability is abysmal in RL research. Lets be real. This needs to be fixed.\n",
        "- TL;DR - Needs moar Flexibility, Stabilitiy, Reproducability \n",
        "\n",
        "#  Framework Features\n",
        "\n",
        "## Agents\n",
        "\n",
        "### DQN\n",
        "\n",
        "![alt text](http://shws.cc.oita-u.ac.jp/shibata/KissSystem.png)\n",
        "\n",
        "###  C51\n",
        "- Basically it is a good old Deep-Q network. But instead of approximating expected future reward values it generates the whole distributions over possible outcomes.\n",
        "-They’ve done it by replacing the output of DQN with probabilities over categorical variables representing different reward values ranges.\n",
        "\n",
        "### Rainbow\n",
        "-Six extensions to the DQN algorithm \n",
        "\n",
        "![alt text](https://ai2-s2-public.s3.amazonaws.com/figures/2017-08-08/0ab3f7ecbdc5a33565a234215604a6ca9d155a33/6-Figure3-1.png)\n",
        "\n",
        "### Implicit Quantile (Very New)\n",
        "\n",
        "![alt text](https://ai2-s2-public.s3.amazonaws.com/figures/2017-08-08/57f85af87e42f1a3d6e74d85a386073049f6586f/4-Figure1-1.png)\n",
        "\n",
        "- Approximates the full quantile function for the state-action return distribution\n",
        "\n",
        "## Atari (Environment, wraps OpenAI's gym)\n",
        "\n",
        "- 60 Atari Environments at https://gym.openai.com/envs/#atari \n",
        "\n",
        "## Common (logging, checkpointing)\n",
        "\n",
        "- Dopamine will save an experiment checkpoint every iteration: one training and one evaluation phase\n",
        "- Checkpoint 1: Experiment statistics (number of iterations performed, learning curves, etc.)\n",
        "- Checkpoint 2: Agent variables, including the tensorflow graph.\n",
        "- Checkpoint 3: Replay buffer data. Atari 2600 replay buffers have a large memory footprint, this prevents that\n",
        "-It also records the agent's performance, both during training and (if enabled) during an optional evaluation phase. \n",
        "\n",
        "## Replay Memory (several neatly wrapped functions)\n",
        "\n",
        "- Circular Replay\n",
        "- Prioritized Replay\n",
        "- Sum Tree\n",
        "\n",
        "## Tests\n",
        "\n",
        "- Self explanatory\n",
        "\n",
        "# Markov Decision Processes\n",
        "\n",
        "![alt text](https://image.slidesharecdn.com/ml-sep-09-091009141615-phpapp01/95/regretbased-reward-elicitation-for-markov-decision-processes-39-728.jpg?cb=1255098159)\n",
        "\n",
        "![alt text](https://i.stack.imgur.com/QNV7Z.jpg)\n",
        "\n",
        "# Policies\n",
        "\n",
        "![alt text](https://image.slidesharecdn.com/deepreinforcementlearningfromscratch1-171221153614/95/deep-reinforcement-learning-from-scratch-29-638.jpg?cb=1513870627)\n",
        "\n",
        "- A policy describes a way of acting. \n",
        "- It is a function that takes in a state and an action and returns the probability of taking that action in that state.\n",
        "-Our goal in reinforcement learning is to learn an optimal policy\n",
        "- An optimal policy is a policy which tells us how to act to maximize return in every state.\n",
        "\n",
        "# Value Functions\n",
        "\n",
        "![alt text](https://slideplayer.com/slide/4757729/15/images/10/Grid+world+Optimal+policy+when+R%28s%29+%3D+for+every+non-terminal+state.jpg)\n",
        "\n",
        "- To learn the optimal policy, we make use of value functions. \n",
        "- There are two types of value functions that are used in reinforcement learning: \n",
        "- 2 types: the state value function, denoted V(s), and the action value function, denoted Q(s, a).\n",
        "- The state value function describes the value of a state when following a policy. It is the expected return when starting from state s acting according to our policy \n",
        "\n",
        "![alt text](https://joshgreaves.com/wp-content/ql-cache/quicklatex.com-814f9fadd2ab8ee9cb85e12999a17eec_l3.svg)\n",
        "\n",
        "- The action value function tells us the value of taking an action in some state when following a certain policy. It is the expected return given the state and action \n",
        "\n",
        "![alt text](https://joshgreaves.com/wp-content/ql-cache/quicklatex.com-5171067fa940af561a4eebe7d3c2d190_l3.svg)\n",
        "\n",
        "# The Bellman Equations\n",
        "\n",
        "- There are 4 Bellman Equations, lets discuss just 2\n",
        "- Theres a Bellman Equation for the state value function and one for the action value function\n",
        "- The Bellman equations let us express values of states as values of other states. \n",
        "-This means that if we know the value of s_{t+1}, we can very easily calculate the value of s_t. \n",
        "- This opens a lot of doors for iterative approaches for calculating the value for each state\n",
        "- Since if we know the value of the next state, we can know the value of the current state.\n",
        "- The Bellman Equation characterizes the optimal values, dynamic programming (+ other methods) help compute them\n",
        "\n",
        "![alt text](https://images.slideplayer.com/16/4958328/slides/slide_18.jpg)"
      ]
    },
    {
      "metadata": {
        "id": "WRNPgSEHt02V",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 510
        },
        "outputId": "10517953-494a-4cde-b9d1-d105f8cd299a"
      },
      "cell_type": "code",
      "source": [
        "# @title Install necessary packages.\n",
        "#dopamine for RL\n",
        "!pip install --upgrade --no-cache-dir dopamine-rl\n",
        "# dopamine dependencies\n",
        "!pip install cmake\n",
        "#Arcade Learning Environment\n",
        "!pip install atari_py"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already up-to-date: dopamine-rl in /usr/local/lib/python3.6/dist-packages (1.0.2)\n",
            "Requirement already satisfied, skipping upgrade: gin-config>=0.1.1 in /usr/local/lib/python3.6/dist-packages (from dopamine-rl) (0.1.1.2)\n",
            "Requirement already satisfied, skipping upgrade: tensorflow in /usr/local/lib/python3.6/dist-packages (from dopamine-rl) (1.10.1)\n",
            "Requirement already satisfied, skipping upgrade: gym>=0.10.5 in /usr/local/lib/python3.6/dist-packages (from dopamine-rl) (0.10.5)\n",
            "Requirement already satisfied, skipping upgrade: opencv-python>=3.4.1.15 in /usr/local/lib/python3.6/dist-packages (from dopamine-rl) (3.4.3.18)\n",
            "Requirement already satisfied, skipping upgrade: absl-py>=0.2.2 in /usr/local/lib/python3.6/dist-packages (from dopamine-rl) (0.4.1)\n",
            "Requirement already satisfied, skipping upgrade: six>=1.10.0 in /usr/local/lib/python3.6/dist-packages (from gin-config>=0.1.1->dopamine-rl) (1.11.0)\n",
            "Requirement already satisfied, skipping upgrade: astor>=0.6.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow->dopamine-rl) (0.7.1)\n",
            "Requirement already satisfied, skipping upgrade: numpy<=1.14.5,>=1.13.3 in /usr/local/lib/python3.6/dist-packages (from tensorflow->dopamine-rl) (1.14.5)\n",
            "Requirement already satisfied, skipping upgrade: gast>=0.2.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow->dopamine-rl) (0.2.0)\n",
            "Requirement already satisfied, skipping upgrade: protobuf>=3.6.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow->dopamine-rl) (3.6.1)\n",
            "Requirement already satisfied, skipping upgrade: tensorboard<1.11.0,>=1.10.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow->dopamine-rl) (1.10.0)\n",
            "Requirement already satisfied, skipping upgrade: termcolor>=1.1.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow->dopamine-rl) (1.1.0)\n",
            "Requirement already satisfied, skipping upgrade: grpcio>=1.8.6 in /usr/local/lib/python3.6/dist-packages (from tensorflow->dopamine-rl) (1.15.0)\n",
            "Requirement already satisfied, skipping upgrade: wheel>=0.26 in /usr/local/lib/python3.6/dist-packages (from tensorflow->dopamine-rl) (0.31.1)\n",
            "Requirement already satisfied, skipping upgrade: setuptools<=39.1.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow->dopamine-rl) (39.1.0)\n",
            "Requirement already satisfied, skipping upgrade: requests>=2.0 in /usr/local/lib/python3.6/dist-packages (from gym>=0.10.5->dopamine-rl) (2.18.4)\n",
            "Requirement already satisfied, skipping upgrade: pyglet>=1.2.0 in /usr/local/lib/python3.6/dist-packages (from gym>=0.10.5->dopamine-rl) (1.3.2)\n",
            "Requirement already satisfied, skipping upgrade: werkzeug>=0.11.10 in /usr/local/lib/python3.6/dist-packages (from tensorboard<1.11.0,>=1.10.0->tensorflow->dopamine-rl) (0.14.1)\n",
            "Requirement already satisfied, skipping upgrade: markdown>=2.6.8 in /usr/local/lib/python3.6/dist-packages (from tensorboard<1.11.0,>=1.10.0->tensorflow->dopamine-rl) (2.6.11)\n",
            "Requirement already satisfied, skipping upgrade: chardet<3.1.0,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests>=2.0->gym>=0.10.5->dopamine-rl) (3.0.4)\n",
            "Requirement already satisfied, skipping upgrade: certifi>=2017.4.17 in /usr/local/lib/python3.6/dist-packages (from requests>=2.0->gym>=0.10.5->dopamine-rl) (2018.8.24)\n",
            "Requirement already satisfied, skipping upgrade: urllib3<1.23,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from requests>=2.0->gym>=0.10.5->dopamine-rl) (1.22)\n",
            "Requirement already satisfied, skipping upgrade: idna<2.7,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests>=2.0->gym>=0.10.5->dopamine-rl) (2.6)\n",
            "Requirement already satisfied, skipping upgrade: future in /usr/local/lib/python3.6/dist-packages (from pyglet>=1.2.0->gym>=0.10.5->dopamine-rl) (0.16.0)\n",
            "Requirement already satisfied: cmake in /usr/local/lib/python3.6/dist-packages (3.12.0)\n",
            "Requirement already satisfied: atari_py in /usr/local/lib/python3.6/dist-packages (0.1.1)\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python3.6/dist-packages (from atari_py) (1.14.5)\n",
            "Requirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from atari_py) (1.11.0)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "Atm0qogwttgW",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# @title Necessary imports and globals.\n",
        "#matrix math\n",
        "import numpy as np\n",
        "#load files\n",
        "import os\n",
        "#DQN for baselines\n",
        "from dopamine.agents.dqn import dqn_agent\n",
        "#high level agent-environment excecution engine\n",
        "from dopamine.atari import run_experiment\n",
        "#visualization + data downloading\n",
        "from dopamine.colab import utils as colab_utils\n",
        "#warnings\n",
        "from absl import flags\n",
        "\n",
        "#where to store training logs\n",
        "BASE_PATH = '/tmp/colab_dope_run'  # @param\n",
        "#which arcade environment?\n",
        "GAME = 'Asterix'  # @param"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "FDbx7hFduBJ9",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# @title Create a  new agent from scratch.\n",
        "\n",
        "#define where to store log data\n",
        "LOG_PATH = os.path.join(BASE_PATH, 'basic_agent', GAME)\n",
        "\n",
        "class BasicAgent(object):\n",
        "  \"\"\"This agent randomly selects an action and sticks to it. It will change\n",
        "  actions with probability switch_prob.\"\"\"\n",
        "  def __init__(self, sess, num_actions, switch_prob=0.1):\n",
        "    #tensorflow session\n",
        "    self._sess = sess\n",
        "    #how many possible actions can it take?\n",
        "    self._num_actions = num_actions\n",
        "    # probability of switching actions in the next timestep?\n",
        "    self._switch_prob = switch_prob\n",
        "    #initialize the action to take (randomly)\n",
        "    self._last_action = np.random.randint(num_actions)\n",
        "    #not debugging\n",
        "    self.eval_mode = False\n",
        "  \n",
        "  #How select an action? \n",
        "  #we define our policy here\n",
        "  def _choose_action(self):\n",
        "    if np.random.random() <= self._switch_prob:\n",
        "      self._last_action = np.random.randint(self._num_actions)\n",
        "    return self._last_action\n",
        "    \n",
        "  #when it checkpoints during training, anything we should do?\n",
        "  def bundle_and_checkpoint(self, unused_checkpoint_dir, unused_iteration):\n",
        "    pass\n",
        "    \n",
        "  #loading from checkpoint\n",
        "  def unbundle(self, unused_checkpoint_dir, unused_checkpoint_version,\n",
        "               unused_data):\n",
        "    pass\n",
        "  \n",
        "  #first action to take\n",
        "  def begin_episode(self, unused_observation):\n",
        "    return self._choose_action()\n",
        "  \n",
        "  #cleanup\n",
        "  def end_episode(self, unused_reward):\n",
        "    pass\n",
        "  \n",
        "  #we can update our policy here\n",
        "  #using the reward and observation\n",
        "  #dynamic programming, Q learning, monte carlo methods, etc.\n",
        "  def step(self, reward, observation):\n",
        "    return self._choose_action()\n",
        "  \n",
        "def create_basic_agent(sess, environment):\n",
        "  \"\"\"The Runner class will expect a function of this type to create an agent.\"\"\"\n",
        "  return BasicAgent(sess, num_actions=environment.action_space.n,\n",
        "                     switch_prob=0.2)\n",
        "\n",
        "# Create the runner class with this agent. We use very small numbers of steps\n",
        "# to terminate quickly, as this is mostly meant for demonstrating how one can\n",
        "# use the framework. We also explicitly terminate after 110 iterations (instead\n",
        "# of the standard 200) to demonstrate the plotting of partial runs.\n",
        "basic_runner = run_experiment.Runner(LOG_PATH,\n",
        "                                      create_basic_agent,\n",
        "                                      game_name=GAME,\n",
        "                                      num_iterations=200,\n",
        "                                      training_steps=10,\n",
        "                                      evaluation_steps=10,\n",
        "                                      max_steps_per_episode=100)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "EYACjtrtuEgt",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 13668
        },
        "outputId": "f416de1e-d7a8-4883-b565-def9f40dde25"
      },
      "cell_type": "code",
      "source": [
        "# @title Train Basic Agent.\n",
        "print('Will train basic agent, please be patient, may be a while...')\n",
        "basic_runner.run_experiment()\n",
        "print('Done training!')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Will train basic agent, please be patient, may be a while...\n",
            "INFO:tensorflow:Beginning training...\n",
            "INFO:tensorflow:Starting iteration 0\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 701.25\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 300.00\n",
            "INFO:tensorflow:Starting iteration 1\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 630.17\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 2\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 686.61\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 3\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 665.08\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 4\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 682.00\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 5\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 637.13\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 6\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 653.94\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 7\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 692.27\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 8\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 690.50\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 9\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 649.99\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 10\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 680.44\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 11\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 684.43\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 0.00\n",
            "INFO:tensorflow:Starting iteration 12\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 689.56\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 13\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 300.00\n",
            "INFO:tensorflow:Average training steps per second: 689.81\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 14\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 693.27\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 15\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 704.10\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 16\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 250.00\n",
            "INFO:tensorflow:Average training steps per second: 704.77\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 17\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 689.82\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 18\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 300.00\n",
            "INFO:tensorflow:Average training steps per second: 678.29\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 19\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 696.14\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 20\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 250.00\n",
            "INFO:tensorflow:Average training steps per second: 694.05\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 21\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 697.97\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 22\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 703.32\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 23\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 700.50\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 24\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 683.18\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 25\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 674.04\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 26\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 703.28\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 27\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 250.00\n",
            "INFO:tensorflow:Average training steps per second: 680.96\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 28\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 689.14\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 29\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 674.35\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 30\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 651.27\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 31\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 689.15\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 250.00\n",
            "INFO:tensorflow:Starting iteration 32\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 686.74\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 33\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 683.00\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 34\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 606.94\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 35\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 687.71\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 250.00\n",
            "INFO:tensorflow:Starting iteration 36\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 637.37\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 0.00\n",
            "INFO:tensorflow:Starting iteration 37\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 250.00\n",
            "INFO:tensorflow:Average training steps per second: 682.30\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 38\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 638.88\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 39\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 672.05\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 40\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 678.76\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 41\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 690.73\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 42\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 602.73\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 43\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 702.99\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 44\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 659.56\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 45\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 701.59\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 46\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 665.49\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 47\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 707.95\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 48\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 707.77\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 49\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 688.20\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 50\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 682.46\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 51\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 704.24\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 52\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 350.00\n",
            "INFO:tensorflow:Average training steps per second: 678.79\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 53\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 676.30\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 54\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 678.84\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 55\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 695.44\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 56\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 685.16\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 57\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 696.06\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 300.00\n",
            "INFO:tensorflow:Starting iteration 58\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 700.33\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 0.00\n",
            "INFO:tensorflow:Starting iteration 59\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 697.73\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 60\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 678.87\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 61\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 685.94\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 62\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 702.18\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 63\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 675.06\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 64\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 684.47\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 65\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 702.89\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 66\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 683.71\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 67\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 635.45\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 68\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 697.92\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 69\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 695.01\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 70\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 668.12\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 71\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 656.17\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 72\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 699.34\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 73\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 697.54\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 74\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 694.26\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 75\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 300.00\n",
            "INFO:tensorflow:Average training steps per second: 640.23\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 76\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 0.00\n",
            "INFO:tensorflow:Average training steps per second: 724.49\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 77\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 670.89\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 78\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 0.00\n",
            "INFO:tensorflow:Average training steps per second: 685.40\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 79\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 674.69\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 80\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 707.61\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 81\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 683.94\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 82\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 699.84\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 300.00\n",
            "INFO:tensorflow:Starting iteration 83\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 703.43\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 84\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 677.52\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 85\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 692.56\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 86\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 685.72\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 87\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 693.77\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 88\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 687.97\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 89\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 250.00\n",
            "INFO:tensorflow:Average training steps per second: 695.78\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 90\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 690.51\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 91\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 717.64\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 92\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 250.00\n",
            "INFO:tensorflow:Average training steps per second: 687.17\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 93\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 710.44\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 94\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 706.07\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 95\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 704.89\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 250.00\n",
            "INFO:tensorflow:Starting iteration 96\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 634.05\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 97\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 701.20\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 98\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 701.36\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 99\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 250.00\n",
            "INFO:tensorflow:Average training steps per second: 709.37\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 100\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 634.40\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 101\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 683.60\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 102\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 696.31\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 103\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 697.45\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 104\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 650.25\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 105\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 677.85\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 106\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 694.23\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 107\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 250.00\n",
            "INFO:tensorflow:Average training steps per second: 667.94\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 108\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 636.76\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 109\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 694.89\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 110\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 695.54\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 111\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 694.71\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 112\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 250.00\n",
            "INFO:tensorflow:Average training steps per second: 688.36\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 113\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 683.01\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 114\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 687.76\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 250.00\n",
            "INFO:tensorflow:Starting iteration 115\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 250.00\n",
            "INFO:tensorflow:Average training steps per second: 683.58\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 116\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 683.59\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 117\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 694.34\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 118\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 692.47\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 119\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 710.14\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 120\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 695.80\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 121\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 696.99\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 122\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 250.00\n",
            "INFO:tensorflow:Average training steps per second: 691.21\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 300.00\n",
            "INFO:tensorflow:Starting iteration 123\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 692.46\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 0.00\n",
            "INFO:tensorflow:Starting iteration 124\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 689.59\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 125\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 666.00\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 250.00\n",
            "INFO:tensorflow:Starting iteration 126\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 687.99\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 127\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 694.15\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 350.00\n",
            "INFO:tensorflow:Starting iteration 128\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 678.25\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 129\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 644.15\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 130\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 705.78\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 131\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 706.92\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 132\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 697.76\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 133\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 649.10\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 134\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 712.50\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 135\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 703.48\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 136\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 705.41\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 137\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 705.05\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 0.00\n",
            "INFO:tensorflow:Starting iteration 138\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 686.42\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 139\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 250.00\n",
            "INFO:tensorflow:Average training steps per second: 690.24\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 140\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 710.83\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 141\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 705.10\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 142\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 721.20\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 143\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 350.00\n",
            "INFO:tensorflow:Average training steps per second: 703.95\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 144\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 718.66\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 145\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 686.21\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 146\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 678.95\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 147\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 659.03\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 250.00\n",
            "INFO:tensorflow:Starting iteration 148\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 703.70\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 149\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 719.18\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 150\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 695.29\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 151\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 698.35\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 152\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 300.00\n",
            "INFO:tensorflow:Average training steps per second: 680.32\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 0.00\n",
            "INFO:tensorflow:Starting iteration 153\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 680.49\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 154\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 655.41\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 155\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 698.43\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 300.00\n",
            "INFO:tensorflow:Starting iteration 156\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 706.68\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 157\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 688.33\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 158\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 621.77\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 159\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 704.64\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 160\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 692.97\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 161\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 680.19\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 162\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 633.29\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 163\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 692.76\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 164\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 699.64\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 165\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 687.45\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 166\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 666.37\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 0.00\n",
            "INFO:tensorflow:Starting iteration 167\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 681.65\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 168\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 350.00\n",
            "INFO:tensorflow:Average training steps per second: 685.29\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 169\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 0.00\n",
            "INFO:tensorflow:Average training steps per second: 683.42\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 170\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 670.91\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 171\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 694.38\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 172\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 250.00\n",
            "INFO:tensorflow:Average training steps per second: 696.27\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 173\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 692.75\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 0.00\n",
            "INFO:tensorflow:Starting iteration 174\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 691.49\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 175\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 682.98\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 176\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 692.31\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 177\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 250.00\n",
            "INFO:tensorflow:Average training steps per second: 699.65\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 178\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 697.08\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 179\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 695.82\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 180\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 696.80\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 181\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 705.17\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 182\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 692.95\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 183\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 664.43\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 200.00\n",
            "INFO:tensorflow:Starting iteration 184\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 690.79\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 185\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 711.73\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 350.00\n",
            "INFO:tensorflow:Starting iteration 186\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 698.71\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 187\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 643.77\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 188\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 0.00\n",
            "INFO:tensorflow:Average training steps per second: 695.66\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 189\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 695.83\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 190\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 693.09\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 191\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 646.50\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 192\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 50.00\n",
            "INFO:tensorflow:Average training steps per second: 703.18\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 250.00\n",
            "INFO:tensorflow:Starting iteration 193\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 678.03\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 194\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 0.00\n",
            "INFO:tensorflow:Average training steps per second: 696.31\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 50.00\n",
            "INFO:tensorflow:Starting iteration 195\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 648.36\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 196\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 100.00\n",
            "INFO:tensorflow:Average training steps per second: 678.81\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "INFO:tensorflow:Starting iteration 197\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 150.00\n",
            "INFO:tensorflow:Average training steps per second: 684.56\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 198\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 691.48\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 100.00\n",
            "INFO:tensorflow:Starting iteration 199\n",
            "INFO:tensorflow:Average undiscounted return per training episode: 200.00\n",
            "INFO:tensorflow:Average training steps per second: 692.28\n",
            "INFO:tensorflow:Average undiscounted return per evaluation episode: 150.00\n",
            "Done training!\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "gzwHF0k0vetd",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# @title Load baseline data\n",
        "!gsutil -q -m cp -R gs://download-dopamine-rl/preprocessed-benchmarks/* /content/\n",
        "experimental_data = colab_utils.load_baselines('/content')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Njtxlx3auIXW",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "fec4129c-263f-4f35-8424-8d0aa3b04708"
      },
      "cell_type": "code",
      "source": [
        "# @title Load the training logs.\n",
        "basic_data = colab_utils.read_experiment(log_path=LOG_PATH, verbose=True)\n",
        "basic_data['agent'] = 'BasicAgent'\n",
        "basic_data['run_number'] = 1\n",
        "experimental_data[GAME] = experimental_data[GAME].merge(basic_data,\n",
        "                                                        how='outer')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Reading statistics from: /tmp/colab_dope_run/sticky_agent/Asterix//logs/log_199\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "G556kKh1uKWN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 512
        },
        "outputId": "4c838e06-3e2b-47b7-9b69-dc33275c11c4"
      },
      "cell_type": "code",
      "source": [
        "# @title Plot training results.\n",
        "\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "fig, ax = plt.subplots(figsize=(16,8))\n",
        "sns.tsplot(data=experimental_data[GAME], time='iteration', unit='run_number',\n",
        "           condition='agent', value='train_episode_returns', ax=ax)\n",
        "plt.title(GAME)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA7wAAAHvCAYAAAB3xtNHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7f63bcbc1ac8>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}