awarebayes/RecNN

View on GitHub
examples/1. Vanilla RL/2. DDPG.ipynb

Summary

Maintainability
Test Coverage
{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DDPG"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Note on this tutorials:**\n",
    "**They mostly contain low level implementations explaining what is going on inside the library.**\n",
    "\n",
    "**Most of the stuff explained here is already available out of the box for your usage.**\n",
    "\n",
    "If you do not care about the detailed implementation with code, go to the [Library Basics]/algorithms how to/ddpg, there is a 20 liner version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/ssd/tools/anaconda3/envs/RecNN/lib/python3.8/site-packages/tqdm/std.py:699: FutureWarning: The Panel class is removed from pandas. Accessing it from the top-level namespace will also be removed in the next version\n",
      "  from pandas import Panel\n",
      "/ssd/tools/anaconda3/envs/RecNN/lib/python3.8/site-packages/tqdm/std.py:699: FutureWarning: The Panel class is removed from pandas. Accessing it from the top-level namespace will also be removed in the next version\n",
      "  from pandas import Panel\n"
     ]
    }
   ],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "from torch.utils.tensorboard import SummaryWriter\n",
    "import torch.nn.functional as F\n",
    "import torch_optimizer as optim\n",
    "\n",
    "from tqdm.auto import tqdm\n",
    "\n",
    "from IPython.display import clear_output\n",
    "%matplotlib inline\n",
    "\n",
    "\n",
    "# == recnn ==\n",
    "import sys\n",
    "sys.path.append(\"../../\")\n",
    "import recnn\n",
    "\n",
    "cuda = torch.device('cuda')\n",
    "\n",
    "# ---\n",
    "frame_size = 10\n",
    "batch_size = 25\n",
    "n_epochs   = 100\n",
    "plot_every = 30\n",
    "step       = 0\n",
    "# --- \n",
    "\n",
    "tqdm.pandas()\n",
    "\n",
    "from jupyterthemes import jtplot\n",
    "jtplot.style(theme='grade3')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# embeddgings: https://drive.google.com/open?id=1EQ_zXBR3DKpmJR3jBgLvt-xoOvArGMsL\n",
    "dirs = recnn.data.env.DataPath(\n",
    "    base=\"../../data/\",\n",
    "    embeddings=\"embeddings/ml20_pca128.pkl\",\n",
    "    ratings=\"ml-20m/ratings.csv\",\n",
    "    cache=\"cache/frame_env.pkl\", # cache will generate after you run\n",
    "    use_cache=True\n",
    ")\n",
    "env = recnn.data.env.FrameEnv(dirs, frame_size, batch_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Actor(nn.Module):\n",
    "    def __init__(self, input_dim, action_dim, hidden_size, init_w=3e-1):\n",
    "        super(Actor, self).__init__()\n",
    "        \n",
    "        self.drop_layer = nn.Dropout(p=0.5)\n",
    "        \n",
    "        self.linear1 = nn.Linear(input_dim, hidden_size)\n",
    "        self.linear2 = nn.Linear(hidden_size, hidden_size)\n",
    "        self.linear3 = nn.Linear(hidden_size, action_dim)\n",
    "        \n",
    "        self.linear3.weight.data.uniform_(-init_w, init_w)\n",
    "        self.linear3.bias.data.uniform_(-init_w, init_w)\n",
    "        \n",
    "    def forward(self, state):\n",
    "        # state = self.state_rep(state)\n",
    "        x = F.relu(self.linear1(state))\n",
    "        x = self.drop_layer(x)\n",
    "        x = F.relu(self.linear2(x))\n",
    "        x = self.drop_layer(x)\n",
    "        # x = torch.tanh(self.linear3(x)) # in case embeds are -1 1 normalized\n",
    "        x = self.linear3(x) # in case embeds are standard scaled / wiped using PCA whitening\n",
    "        # return state, x\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Critic(nn.Module):\n",
    "    def __init__(self, input_dim, action_dim, hidden_size, init_w=3e-5):\n",
    "        super(Critic, self).__init__()\n",
    "        \n",
    "        self.drop_layer = nn.Dropout(p=0.5)\n",
    "        \n",
    "        self.linear1 = nn.Linear(input_dim + action_dim, hidden_size)\n",
    "        self.linear2 = nn.Linear(hidden_size, hidden_size)\n",
    "        self.linear3 = nn.Linear(hidden_size, 1)\n",
    "        \n",
    "        self.linear3.weight.data.uniform_(-init_w, init_w)\n",
    "        self.linear3.bias.data.uniform_(-init_w, init_w)\n",
    "        \n",
    "    def forward(self, state, action):\n",
    "        x = torch.cat([state, action], 1)\n",
    "        x = F.relu(self.linear1(x))\n",
    "        x = self.drop_layer(x)\n",
    "        x = F.relu(self.linear2(x))\n",
    "        x = self.drop_layer(x)\n",
    "        x = self.linear3(x)\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def soft_update(net, target_net, soft_tau=1e-2):\n",
    "    for target_param, param in zip(target_net.parameters(), net.parameters()):\n",
    "            target_param.data.copy_(\n",
    "                target_param.data * (1.0 - soft_tau) + param.data * soft_tau\n",
    "            )\n",
    "            \n",
    "def run_tests():\n",
    "    test_batch = next(iter(env.test_dataloader))\n",
    "    losses = ddpg_update(test_batch, params, learn=False, step=step)\n",
    "    \n",
    "    gen_actions = debug['next_action']\n",
    "    true_actions = env.base.embeddings.detach().cpu().numpy()\n",
    "    \n",
    "    f = plotter.kde_reconstruction_error(ad, gen_actions, true_actions, cuda)\n",
    "    writer.add_figure('rec_error',f, losses['step'])\n",
    "    return losses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def ddpg_update(batch, params, learn=True, step=-1):\n",
    "    \n",
    "    state, action, reward, next_state, done = recnn.data.get_base_batch(batch)\n",
    "        \n",
    "    # --------------------------------------------------------#\n",
    "    # Value Learning\n",
    "    \n",
    "    with torch.no_grad():\n",
    "        next_action = target_policy_net(next_state)\n",
    "        target_value   = target_value_net(next_state, next_action.detach())\n",
    "        expected_value = reward + (1.0 - done) * params['gamma'] * target_value\n",
    "        expected_value = torch.clamp(expected_value,\n",
    "                                     params['min_value'], params['max_value'])\n",
    "\n",
    "    value = value_net(state, action)\n",
    "    \n",
    "    value_loss = torch.pow(value - expected_value.detach(), 2).mean()\n",
    "    \n",
    "    if learn:\n",
    "        value_optimizer.zero_grad()\n",
    "        value_loss.backward()\n",
    "        value_optimizer.step()\n",
    "    else:\n",
    "        debug['next_action'] = next_action\n",
    "        writer.add_figure('next_action',\n",
    "                    recnn.utils.pairwise_distances_fig(next_action[:50]), step)\n",
    "        writer.add_histogram('value', value, step)\n",
    "        writer.add_histogram('target_value', target_value, step)\n",
    "        writer.add_histogram('expected_value', expected_value, step)\n",
    "    \n",
    "    # --------------------------------------------------------#\n",
    "    # Policy learning\n",
    "    \n",
    "    gen_action = policy_net(state)\n",
    "    policy_loss = -value_net(state, gen_action)\n",
    "    \n",
    "    if not learn:\n",
    "        debug['gen_action'] = gen_action\n",
    "        writer.add_histogram('policy_loss', policy_loss, step)\n",
    "        writer.add_figure('next_action',\n",
    "                    recnn.utils.pairwise_distances_fig(gen_action[:50]), step)\n",
    "        \n",
    "    policy_loss = policy_loss.mean()\n",
    "    \n",
    "    if learn and step % params['policy_step']== 0:\n",
    "        policy_optimizer.zero_grad()\n",
    "        policy_loss.backward()\n",
    "        torch.nn.utils.clip_grad_norm_(policy_net.parameters(), -1, 1)\n",
    "        policy_optimizer.step()\n",
    "\n",
    "        soft_update(value_net, target_value_net, soft_tau=params['soft_tau'])\n",
    "        soft_update(policy_net, target_policy_net, soft_tau=params['soft_tau'])\n",
    "\n",
    "\n",
    "    losses = {'value': value_loss.item(), 'policy': policy_loss.item(), 'step': step}\n",
    "    recnn.utils.write_losses(writer, losses, kind='train' if learn else 'test')\n",
    "    return losses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# === ddpg settings ===\n",
    "\n",
    "params = {\n",
    "    'gamma'      : 0.99,\n",
    "    'min_value'  : -10,\n",
    "    'max_value'  : 10,\n",
    "    'policy_step': 10,\n",
    "    'soft_tau'   : 0.001,\n",
    "    \n",
    "    'policy_lr'  : 1e-5,\n",
    "    'value_lr'   : 1e-5,\n",
    "    'actor_weight_init': 54e-2,\n",
    "    'critic_weight_init': 6e-1,\n",
    "}\n",
    "\n",
    "# === end ==="
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "value_net  = Critic(1290, 128, 256, params['critic_weight_init']).to(cuda)\n",
    "policy_net = Actor(1290, 128, 256, params['actor_weight_init']).to(cuda)\n",
    "\n",
    "\n",
    "target_value_net = Critic(1290, 128, 256).to(cuda)\n",
    "target_policy_net = Actor(1290, 128, 256).to(cuda)\n",
    "\n",
    "ad = recnn.nn.models.AnomalyDetector().to(cuda)\n",
    "ad.load_state_dict(torch.load('../../models/anomaly.pt'))\n",
    "ad.eval()\n",
    "\n",
    "target_policy_net.eval()\n",
    "target_value_net.eval()\n",
    "\n",
    "soft_update(value_net, target_value_net, soft_tau=1.0)\n",
    "soft_update(policy_net, target_policy_net, soft_tau=1.0)\n",
    "\n",
    "value_criterion = nn.MSELoss()\n",
    "\n",
    "# from good to bad: Ranger Radam Adam RMSprop\n",
    "value_optimizer = optim.Ranger(value_net.parameters(),\n",
    "                              lr=params['value_lr'], weight_decay=1e-2)\n",
    "policy_optimizer = optim.Ranger(policy_net.parameters(),\n",
    "                               lr=params['policy_lr'], weight_decay=1e-5)\n",
    "\n",
    "loss = {\n",
    "    'test': {'value': [], 'policy': [], 'step': []},\n",
    "    'train': {'value': [], 'policy': [], 'step': []}\n",
    "    }\n",
    "\n",
    "debug = {}\n",
    "\n",
    "writer = SummaryWriter(log_dir='../../runs')\n",
    "plotter = recnn.utils.Plotter(loss, [['value', 'policy']],)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "step 990\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "../../recnn/utils/plot.py:101: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).\n",
      "  ad.rec_error(torch.tensor(actions).to(device).float())\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x432 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "ename": "AssertionError",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAssertionError\u001b[0m                            Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-9-31a90ad27f2b>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m     11\u001b[0m             \u001b[0mplotter\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot_loss\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     12\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mstep\u001b[0m \u001b[0;34m>\u001b[0m \u001b[0;36m1000\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 13\u001b[0;31m             \u001b[0;32massert\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mAssertionError\u001b[0m: "
     ]
    }
   ],
   "source": [
    "for epoch in range(n_epochs):\n",
    "    for batch in tqdm(env.train_dataloader):\n",
    "        loss = ddpg_update(batch, params, step=step)\n",
    "        plotter.log_losses(loss)\n",
    "        step += 1\n",
    "        if step % plot_every == 0:\n",
    "            clear_output(True)\n",
    "            print('step', step)\n",
    "            test_loss = run_tests()\n",
    "            plotter.log_losses(test_loss, test=True)\n",
    "            plotter.plot_loss()\n",
    "        if step > 1000:\n",
    "            assert False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "torch.save(value_net.state_dict(), \"../../models/ddpg_value.pt\")\n",
    "torch.save(policy_net.state_dict(), \"../../models/ddpg_policy.pt\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Reconstruction error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "../../recnn/utils/plot.py:101: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).\n",
      "  ad.rec_error(torch.tensor(actions).to(device).float())\n",
      "../../recnn/utils/plot.py:126: UserWarning: Matplotlib is currently using module://ipykernel.pylab.backend_inline, which is a non-GUI backend, so cannot show the figure.\n",
      "  fig.show()\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "gen_actions = debug['next_action']\n",
    "true_actions = env.base.embeddings.numpy()\n",
    "\n",
    "\n",
    "ad = recnn.nn.AnomalyDetector().to(cuda)\n",
    "ad.load_state_dict(torch.load('../../models/anomaly.pt'))\n",
    "ad.eval()\n",
    "\n",
    "plotter.plot_kde_reconstruction_error(ad, gen_actions, true_actions, cuda)"
   ]
  },
  {
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}