14.PPO.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Proximal Policy Optimization"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import gym\n",
    "gym.logger.set_level(40)\n",
    "\n",
    "import numpy as np\n",
    "\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import torch.optim as optim \n",
    "from torch.utils.data.sampler import BatchSampler, SubsetRandomSampler\n",
    "\n",
    "from IPython.display import clear_output\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from timeit import default_timer as timer\n",
    "from datetime import timedelta\n",
    "import os\n",
    "import glob\n",
    "\n",
    "\n",
    "from baselines.common.vec_env.dummy_vec_env import DummyVecEnv\n",
    "from baselines.common.vec_env.subproc_vec_env import SubprocVecEnv\n",
    "\n",
    "from utils.hyperparameters import Config\n",
    "from utils.plot import plot\n",
    "from utils.wrappers import make_env_a2c_atari\n",
    "from agents.A2C import Model as A2C"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Hyperparameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "log_dir = \"/tmp/gym/\"\n",
    "\n",
    "try:\n",
    "    os.makedirs(log_dir)\n",
    "except OSError:\n",
    "    files = glob.glob(os.path.join(log_dir, '*.monitor.csv'))\n",
    "    for f in files:\n",
    "        os.remove(f)\n",
    "\n",
    "config = Config()\n",
    "\n",
    "#ppo control\n",
    "config.ppo_epoch = 3\n",
    "config.num_mini_batch = 32\n",
    "config.ppo_clip_param = 0.1\n",
    "\n",
    "#a2c control\n",
    "config.num_agents=8\n",
    "config.rollout=128\n",
    "config.USE_GAE = True\n",
    "config.gae_tau = 0.95\n",
    "\n",
    "#misc agent variables\n",
    "config.GAMMA=0.99\n",
    "config.LR=7e-4\n",
    "config.entropy_loss_weight=0.01\n",
    "config.value_loss_weight=1.0\n",
    "config.grad_norm_max = 0.5\n",
    "\n",
    "config.MAX_FRAMES=int(1e7 / config.num_agents / config.rollout)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Rollout Storage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "class RolloutStorage(object):\n",
    "    def __init__(self, num_steps, num_processes, obs_shape, action_space, device, USE_GAE=True, gae_tau=0.95):\n",
    "        self.observations = torch.zeros(num_steps + 1, num_processes, *obs_shape).to(device)\n",
    "        self.rewards = torch.zeros(num_steps, num_processes, 1).to(device)\n",
    "        self.value_preds = torch.zeros(num_steps + 1, num_processes, 1).to(device)\n",
    "        self.returns = torch.zeros(num_steps + 1, num_processes, 1).to(device)\n",
    "        self.action_log_probs = torch.zeros(num_steps, num_processes, 1).to(device)\n",
    "        self.actions = torch.zeros(num_steps, num_processes, 1).to(device, torch.long)\n",
    "        self.masks = torch.ones(num_steps + 1, num_processes, 1).to(device)\n",
    "\n",
    "        self.num_steps = num_steps\n",
    "        self.step = 0\n",
    "        self.gae = USE_GAE\n",
    "        self.gae_tau = gae_tau\n",
    "\n",
    "    def insert(self, current_obs, action, action_log_prob, value_pred, reward, mask):\n",
    "        self.observations[self.step + 1].copy_(current_obs)\n",
    "        self.actions[self.step].copy_(action)\n",
    "        self.action_log_probs[self.step].copy_(action_log_prob)\n",
    "        self.value_preds[self.step].copy_(value_pred)\n",
    "        self.rewards[self.step].copy_(reward)\n",
    "        self.masks[self.step + 1].copy_(mask)\n",
    "\n",
    "        self.step = (self.step + 1) % self.num_steps\n",
    "\n",
    "    def after_update(self):\n",
    "        self.observations[0].copy_(self.observations[-1])\n",
    "        self.masks[0].copy_(self.masks[-1])\n",
    "\n",
    "    def compute_returns(self, next_value, gamma):\n",
    "        if self.gae:\n",
    "            self.value_preds[-1] = next_value\n",
    "            gae = 0\n",
    "            for step in reversed(range(self.rewards.size(0))):\n",
    "                delta = self.rewards[step] + gamma * self.value_preds[step + 1] * self.masks[step + 1] - self.value_preds[step]\n",
    "                gae = delta + gamma * self.gae_tau * self.masks[step + 1] * gae\n",
    "                self.returns[step] = gae + self.value_preds[step]\n",
    "        else:\n",
    "            self.returns[-1] = next_value\n",
    "            for step in reversed(range(self.rewards.size(0))):\n",
    "                self.returns[step] = self.returns[step + 1] * \\\n",
    "                    gamma * self.masks[step + 1] + self.rewards[step]\n",
    "\n",
    "    def feed_forward_generator(self, advantages, num_mini_batch):\n",
    "        num_steps, num_processes = self.rewards.size()[0:2]\n",
    "        batch_size = num_processes * num_steps\n",
    "        assert batch_size >= num_mini_batch, (\n",
    "            f\"PPO requires the number processes ({num_processes}) \"\n",
    "            f\"* number of steps ({num_steps}) = {num_processes * num_steps} \"\n",
    "            f\"to be greater than or equal to the number of PPO mini batches ({num_mini_batch}).\")\n",
    "        mini_batch_size = batch_size // num_mini_batch\n",
    "        sampler = BatchSampler(SubsetRandomSampler(range(batch_size)), mini_batch_size, drop_last=False)\n",
    "        for indices in sampler:\n",
    "            observations_batch = self.observations[:-1].view(-1,\n",
    "                                        *self.observations.size()[2:])[indices]\n",
    "            actions_batch = self.actions.view(-1, self.actions.size(-1))[indices]\n",
    "            return_batch = self.returns[:-1].view(-1, 1)[indices]\n",
    "            masks_batch = self.masks[:-1].view(-1, 1)[indices]\n",
    "            old_action_log_probs_batch = self.action_log_probs.view(-1, 1)[indices]\n",
    "            adv_targ = advantages.view(-1, 1)[indices]\n",
    "\n",
    "            yield observations_batch, actions_batch, return_batch, masks_batch, old_action_log_probs_batch, adv_targ"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Agent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Model(A2C):\n",
    "    def __init__(self, static_policy=False, env=None, config=None):\n",
    "        super(Model, self).__init__(static_policy, env, config)\n",
    "        \n",
    "        self.num_agents = config.num_agents\n",
    "        self.value_loss_weight = config.value_loss_weight\n",
    "        self.entropy_loss_weight = config.entropy_loss_weight\n",
    "        self.rollout = config.rollout\n",
    "        self.grad_norm_max = config.grad_norm_max\n",
    "\n",
    "        self.ppo_epoch = config.ppo_epoch\n",
    "        self.num_mini_batch = config.num_mini_batch\n",
    "        self.clip_param = config.ppo_clip_param\n",
    "\n",
    "        self.optimizer = optim.Adam(self.model.parameters(), lr=self.lr, eps=1e-5)\n",
    "        \n",
    "        self.rollouts = RolloutStorage(self.rollout, self.num_agents,\n",
    "            self.num_feats, self.env.action_space, self.device, config.USE_GAE, config.gae_tau)\n",
    "\n",
    "    def compute_loss(self, sample):\n",
    "        observations_batch, actions_batch, return_batch, masks_batch, old_action_log_probs_batch, adv_targ = sample\n",
    "\n",
    "        values, action_log_probs, dist_entropy = self.evaluate_actions(observations_batch, actions_batch)\n",
    "\n",
    "        ratio = torch.exp(action_log_probs - old_action_log_probs_batch)\n",
    "        surr1 = ratio * adv_targ\n",
    "        surr2 = torch.clamp(ratio, 1.0 - self.clip_param, 1.0 + self.clip_param) * adv_targ\n",
    "        action_loss = -torch.min(surr1, surr2).mean()\n",
    "\n",
    "        value_loss = F.mse_loss(return_batch, values)\n",
    "\n",
    "        loss = action_loss + self.value_loss_weight * value_loss - self.entropy_loss_weight * dist_entropy\n",
    "\n",
    "        return loss, action_loss, value_loss, dist_entropy\n",
    "\n",
    "    def update(self, rollout):\n",
    "        advantages = rollout.returns[:-1] - rollout.value_preds[:-1]\n",
    "        advantages = (advantages - advantages.mean()) / (\n",
    "            advantages.std() + 1e-5)\n",
    "\n",
    "\n",
    "        value_loss_epoch = 0\n",
    "        action_loss_epoch = 0\n",
    "        dist_entropy_epoch = 0\n",
    "\n",
    "        for e in range(self.ppo_epoch):\n",
    "            data_generator = rollout.feed_forward_generator(\n",
    "                advantages, self.num_mini_batch)\n",
    "\n",
    "            for sample in data_generator:\n",
    "                loss, action_loss, value_loss, dist_entropy = self.compute_loss(sample)\n",
    "\n",
    "                self.optimizer.zero_grad()\n",
    "                loss.backward()\n",
    "                torch.nn.utils.clip_grad_norm_(self.model.parameters(), self.grad_norm_max)\n",
    "                self.optimizer.step()\n",
    "\n",
    "                value_loss_epoch += value_loss.item()\n",
    "                action_loss_epoch += action_loss.item()\n",
    "                dist_entropy_epoch += dist_entropy.item()\n",
    "        \n",
    "        value_loss_epoch /= (self.ppo_epoch * self.num_mini_batch)\n",
    "        action_loss_epoch /= (self.ppo_epoch * self.num_mini_batch)\n",
    "        dist_entropy_epoch /= (self.ppo_epoch * self.num_mini_batch)\n",
    "        total_loss = value_loss_epoch + action_loss_epoch + dist_entropy_epoch\n",
    "\n",
    "        #self.save_loss(total_loss, action_loss_epoch, value_loss_epoch, dist_entropy_epoch)\n",
    "\n",
    "        return action_loss_epoch, value_loss_epoch, dist_entropy_epoch"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training Loop"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Updates 9700, Num Timesteps 9933824, FPS 970,\n",
      "Mean/Median Reward 21.0/21.0, Min/Max Reward 21.0/21.0,\n",
      "Entropy 0.90210, Value Loss -0.00484, Policy Loss 0.00291\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "if __name__=='__main__':\n",
    "    seed = 1\n",
    "\n",
    "    torch.manual_seed(seed)\n",
    "    if torch.cuda.is_available():\n",
    "        torch.cuda.manual_seed(seed)\n",
    "\n",
    "    torch.set_num_threads(1)\n",
    "\n",
    "    env_id = \"PongNoFrameskip-v4\"\n",
    "    envs = [make_env_a2c_atari(env_id, seed, i, log_dir) for i in range(config.num_agents)]\n",
    "    envs = SubprocVecEnv(envs) if config.num_agents > 1 else DummyVecEnv(envs)\n",
    "\n",
    "    obs_shape = envs.observation_space.shape\n",
    "    obs_shape = (obs_shape[0] * 4, *obs_shape[1:])\n",
    "\n",
    "    model = Model(env=envs, config=config)\n",
    "\n",
    "    current_obs = torch.zeros(config.num_agents, *obs_shape,\n",
    "                    device=config.device, dtype=torch.float)\n",
    "\n",
    "    def update_current_obs(obs):\n",
    "        shape_dim0 = envs.observation_space.shape[0]\n",
    "        obs = torch.from_numpy(obs.astype(np.float32)).to(config.device)\n",
    "        current_obs[:, :-shape_dim0] = current_obs[:, shape_dim0:]\n",
    "        current_obs[:, -shape_dim0:] = obs\n",
    "\n",
    "    obs = envs.reset()\n",
    "    update_current_obs(obs)\n",
    "\n",
    "    model.rollouts.observations[0].copy_(current_obs)\n",
    "    \n",
    "    episode_rewards = np.zeros(config.num_agents, dtype=np.float)\n",
    "    final_rewards = np.zeros(config.num_agents, dtype=np.float)\n",
    "\n",
    "    start=timer()\n",
    "\n",
    "    print_step = 1\n",
    "    print_threshold = 10\n",
    "    \n",
    "    for frame_idx in range(1, config.MAX_FRAMES+1):\n",
    "        for step in range(config.rollout):\n",
    "            with torch.no_grad():\n",
    "                values, actions, action_log_prob = model.get_action(model.rollouts.observations[step])\n",
    "            cpu_actions = actions.view(-1).cpu().numpy()\n",
    "    \n",
    "            obs, reward, done, _ = envs.step(cpu_actions)\n",
    "\n",
    "            episode_rewards += reward\n",
    "            masks = 1. - done.astype(np.float32)\n",
    "            final_rewards *= masks\n",
    "            final_rewards += (1. - masks) * episode_rewards\n",
    "            episode_rewards *= masks\n",
    "\n",
    "            rewards = torch.from_numpy(reward.astype(np.float32)).view(-1, 1).to(config.device)\n",
    "            masks = torch.from_numpy(masks).to(config.device).view(-1, 1)\n",
    "\n",
    "            current_obs *= masks.view(-1, 1, 1, 1)\n",
    "            update_current_obs(obs)\n",
    "\n",
    "            model.rollouts.insert(current_obs, actions.view(-1, 1), action_log_prob, values, rewards, masks)\n",
    "            \n",
    "        with torch.no_grad():\n",
    "            next_value = model.get_values(model.rollouts.observations[-1])\n",
    "\n",
    "        model.rollouts.compute_returns(next_value, config.GAMMA)\n",
    "            \n",
    "        value_loss, action_loss, dist_entropy = model.update(model.rollouts)\n",
    "        \n",
    "        model.rollouts.after_update()\n",
    "\n",
    "        if frame_idx % 100 == 0:\n",
    "            try:\n",
    "                clear_output()\n",
    "                end = timer()\n",
    "                total_num_steps = (frame_idx + 1) * config.num_agents * config.rollout\n",
    "                print(\"Updates {}, Num Timesteps {}, FPS {},\\nMean/Median Reward {:.1f}/{:.1f}, Min/Max Reward {:.1f}/{:.1f},\\nEntropy {:.5f}, Value Loss {:.5f}, Policy Loss {:.5f}\".\n",
    "                format(frame_idx, total_num_steps,\n",
    "                       int(total_num_steps / (end - start)),\n",
    "                       np.mean(final_rewards),\n",
    "                       np.median(final_rewards),\n",
    "                       np.min(final_rewards),\n",
    "                       np.max(final_rewards), dist_entropy,\n",
    "                       value_loss, action_loss))\n",
    "                plot(log_dir, \"PongNoFrameskip-v4\", 'PPO', \n",
    "                     config.MAX_FRAMES * config.num_agents * config.rollout)\n",
    "            except IOError:\n",
    "                pass\n",
    "\n",
    "    model.save_w()\n",
    "    envs.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}