prior.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "prior.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "metadata": {
        "id": "f5A7LWa0XoLM",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "Part of https://github.com/htdt/cartpole-solved"
      ]
    },
    {
      "metadata": {
        "id": "XB3hFYr6hisD",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "!pip install torch_nightly -f https://download.pytorch.org/whl/nightly/cpu/torch_nightly.html\n",
        "!pip install -q gym ipdb"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "zJr-1s4fYuhB",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "!wget -N -q -o - https://raw.githubusercontent.com/htdt/cartpole-solved/master/segment_tree.py"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "tEIxLMSmZ18R",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "from collections import deque\n",
        "import gym, random, math, ipdb\n",
        "from tqdm import trange\n",
        "import numpy as np\n",
        "\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.optim as optim\n",
        "\n",
        "from segment_tree import SumDeque\n",
        "\n",
        "gym.logger.set_level(40)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "NIuUihaPJw1D",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "GzvYKXl8ig8Q",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "class DQN(nn.Module):\n",
        "  input_dim = 4\n",
        "  output_dim = 2\n",
        "  hidden = 32\n",
        "\n",
        "  def __init__(self):\n",
        "    super(DQN, self).__init__()\n",
        "    self.features = nn.Sequential(\n",
        "      nn.Linear(self.input_dim, self.hidden),\n",
        "      nn.ReLU(),\n",
        "      nn.Linear(self.hidden, self.hidden),\n",
        "      nn.ReLU())\n",
        "    self.adv = nn.Linear(self.hidden, self.output_dim)\n",
        "    self.val = nn.Linear(self.hidden, 1)\n",
        "\n",
        "  def forward(self, x):\n",
        "    x = self.features(x)\n",
        "    adv = self.adv(x)\n",
        "    val = self.val(x)\n",
        "    return val + adv - adv.mean(1, keepdim=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "FZblc6DQjOEe",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "class Agent:\n",
        "  output_dim = 2\n",
        "  lr = 1e-3\n",
        "  batch_size = 32\n",
        "  gamma = 0.99\n",
        "\n",
        "  def __init__(self):\n",
        "    self.online_net = DQN()\n",
        "    self.online_net.train()\n",
        "\n",
        "    self.target_net = DQN()\n",
        "    self.update_target_net()\n",
        "    for param in self.target_net.parameters(): param.requires_grad = False\n",
        "\n",
        "    self.optimiser = optim.Adam(self.online_net.parameters(), lr=self.lr)\n",
        "\n",
        "  def act(self, state):\n",
        "    state = torch.FloatTensor(state).unsqueeze(0)\n",
        "    with torch.no_grad():\n",
        "      return self.online_net(state).argmax(1).item()\n",
        "\n",
        "  def act_e_greedy(self, state, epsilon=0.01):\n",
        "    if random.random() < epsilon:\n",
        "      return random.randrange(self.output_dim)\n",
        "    else:\n",
        "      return self.act(state)\n",
        "\n",
        "  def train_iter(self, buffer):\n",
        "    state, action, reward, next_state, terminal, weights, idx = buffer.sample(self.batch_size)\n",
        "       \n",
        "    q_value = self.online_net(state)[range(self.batch_size), action]\n",
        "    with torch.no_grad():\n",
        "      next_state_action = self.online_net(next_state).max(1)[1] # online_net to get action for next_state\n",
        "      next_qv = self.target_net(next_state)[range(self.batch_size), next_state_action] # target_net to get Q\n",
        "      target_qv = reward + self.gamma * (1 - terminal) * next_qv\n",
        "    \n",
        "    loss = (q_value - target_qv).pow(2)\n",
        "    #loss *= weights\n",
        "    buffer.update_prior(idx, loss.tolist())\n",
        "\n",
        "    self.optimiser.zero_grad()\n",
        "    loss.mean().backward()\n",
        "    self.optimiser.step()\n",
        "\n",
        "\n",
        "  def update_target_net(self):\n",
        "    self.target_net.load_state_dict(self.online_net.state_dict())"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "LREzvbrWks_j",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "class ReplayBuffer():\n",
        "  def __init__(self, capacity):\n",
        "    self.capacity = capacity\n",
        "    self._buffer = deque(maxlen=capacity)\n",
        "    self._prior = SumDeque(capacity)\n",
        "    self._prior.max = 1.\n",
        "  \n",
        "  def append(self, t):\n",
        "    self._buffer.append(t)\n",
        "    self._prior.append(self._prior.max)\n",
        "    \n",
        "  def sample(self, batch_size):\n",
        "    idx = [self._prior.find_prefixsum_idx(random.random() * self._prior.sum()) for _ in range(batch_size)]\n",
        "    weights = 1. / np.sqrt([self._prior[i] for i in idx]) # hardcoded beta = .5\n",
        "    weights /= weights.max()\n",
        "    tt = [self._buffer[i] for i in idx]\n",
        "    s, a, r, ns, t = zip(*tt)\n",
        "    f, l = torch.FloatTensor, torch.LongTensor\n",
        "    return f(s), l(a), f(r), f(ns), f(t), f(weights), idx\n",
        "\n",
        "  def update_prior(self, idxes, losses):\n",
        "    for idx, loss in zip(idxes, losses):\n",
        "      self._prior[idx] = np.sqrt(loss) # hardcoded alpha = .5\n",
        "\n",
        "  def __len__(self):\n",
        "    return len(self._buffer)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "0T0EHr87jQTe",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 355
        },
        "outputId": "6f7cd253-67d6-4b45-d5a5-58150f67820e"
      },
      "cell_type": "code",
      "source": [
        "buffer = ReplayBuffer(5000)\n",
        "env = gym.make(\"CartPole-v0\")\n",
        "agent = Agent()\n",
        "\n",
        "rewards = []\n",
        "episode_reward = 0\n",
        "state = env.reset()\n",
        "\n",
        "for i in trange(10000):\n",
        "  eps = .01 + .99 * math.exp(-i / 500) # close to .01 at 3500\n",
        "  action = agent.act_e_greedy(state, eps)\n",
        "  next_state, reward, done, _ = env.step(action)\n",
        "  buffer.append((state, action, reward, next_state, done))\n",
        "  \n",
        "  state = next_state\n",
        "  episode_reward += reward\n",
        "  \n",
        "  if done:\n",
        "    state = env.reset()\n",
        "    rewards.append(episode_reward)\n",
        "    episode_reward = 0\n",
        "    \n",
        "  if i > 100:\n",
        "    agent.train_iter(buffer)\n",
        "   \n",
        "  if i > 100 and i % 100 == 0:\n",
        "    agent.update_target_net()\n",
        "\n",
        "plt.figure(figsize=(20,5))\n",
        "plt.subplot(131)\n",
        "plt.plot(rewards)\n",
        "plt.subplot(132)\n",
        "plt.plot(list(buffer._prior))\n",
        "plt.plot(list(map(lambda t: int(t[-1]), buffer._buffer)))\n",
        "print(buffer._prior.max)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "100%|██████████| 10000/10000 [01:29<00:00, 176.72it/s]\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "76.87257364057092\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAvgAAAEwCAYAAADCcQWnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7f21e5ce7fd0>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "g2nKbWOtsyjz",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}