PlayTransformer/naive-transformer-time-series-test.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Implement a simple Transformer\n",
    "\n",
    "This notebook implement a simple transformer model, trained on a randomly generated time series data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import mean_squared_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Generate synthetic time series data\n",
    "def generate_time_series(n_samples, n_steps):\n",
    "    freq1, freq2, offset1, offset2 = np.random.rand(4, n_samples, 1)\n",
    "    time = np.linspace(0, 1, n_steps)\n",
    "    series = 0.5 * np.sin((time - offset1) * (freq1 * 10 + 10))\n",
    "    series += 0.2 * np.sin((time - offset2) * (freq2 * 20 + 20))\n",
    "    series += 0.1 * (np.random.rand(n_samples, n_steps) - 0.5)\n",
    "    return series.astype(np.float32)\n",
    "\n",
    "# Create a transformer-based model\n",
    "class TransformerModel(nn.Module):\n",
    "    def __init__(self, input_size, output_size, n_layers, n_heads, hidden_size, dropout):\n",
    "        super(TransformerModel, self).__init__()\n",
    "        self.transformer = nn.Transformer(\n",
    "            d_model=input_size,\n",
    "            nhead=n_heads,\n",
    "            num_encoder_layers=n_layers,\n",
    "            num_decoder_layers=n_layers,\n",
    "            dim_feedforward=hidden_size,\n",
    "            dropout=dropout,\n",
    "        )\n",
    "        self.fc = nn.Linear(input_size, output_size)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = self.transformer(x, x)\n",
    "        x = self.fc(x)\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using device: cuda\n"
     ]
    }
   ],
   "source": [
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "print(f\"Using device: {device}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Prepare data\n",
    "n_samples = 1000\n",
    "n_steps = 513\n",
    "series = generate_time_series(n_samples, n_steps)\n",
    "scaler = StandardScaler()\n",
    "series_scaled = scaler.fit_transform(series.T).T\n",
    "X = series_scaled[:, :n_steps-1]\n",
    "y = series_scaled[:, 1:]\n",
    "\n",
    "# Split data into train and test sets\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)\n",
    "\n",
    "# Convert data to PyTorch tensors\n",
    "X_train_tensor = torch.tensor(X_train, dtype=torch.float32)\n",
    "y_train_tensor = torch.tensor(y_train, dtype=torch.float32)\n",
    "X_test_tensor = torch.tensor(X_test, dtype=torch.float32)\n",
    "y_test_tensor = torch.tensor(y_test, dtype=torch.float32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/root/miniconda3/envs/torch210/lib/python3.11/site-packages/torch/nn/modules/transformer.py:282: UserWarning: enable_nested_tensor is True, but self.use_nested_tensor is False because encoder_layer.self_attn.batch_first was not True(use batch_first for better inference performance)\n",
      "  warnings.warn(f\"enable_nested_tensor is True, but self.use_nested_tensor is False because {why_not_sparsity_fast_path}\")\n",
      "/root/miniconda3/envs/torch210/lib/python3.11/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch [1/200], Loss: 1.2826976776123047\n",
      "Epoch [2/200], Loss: 1.0116225481033325\n",
      "Epoch [3/200], Loss: 0.635532021522522\n",
      "Epoch [4/200], Loss: 0.4489162862300873\n",
      "Epoch [5/200], Loss: 0.35763952136039734\n",
      "Epoch [6/200], Loss: 0.2987120449542999\n",
      "Epoch [7/200], Loss: 0.30284008383750916\n",
      "Epoch [8/200], Loss: 0.2880264222621918\n",
      "Epoch [9/200], Loss: 0.29169321060180664\n",
      "Epoch [10/200], Loss: 0.2669679522514343\n",
      "Epoch [11/200], Loss: 0.2674340307712555\n",
      "Epoch [12/200], Loss: 0.25883418321609497\n",
      "Epoch [13/200], Loss: 0.23866242170333862\n",
      "Epoch [14/200], Loss: 0.21799877285957336\n",
      "Epoch [15/200], Loss: 0.20396709442138672\n",
      "Epoch [16/200], Loss: 0.19301307201385498\n",
      "Epoch [17/200], Loss: 0.17444251477718353\n",
      "Epoch [18/200], Loss: 0.16921290755271912\n",
      "Epoch [19/200], Loss: 0.1575583964586258\n",
      "Epoch [20/200], Loss: 0.14533430337905884\n",
      "Epoch [21/200], Loss: 0.1386234015226364\n",
      "Epoch [22/200], Loss: 0.12826502323150635\n",
      "Epoch [23/200], Loss: 0.11779733747243881\n",
      "Epoch [24/200], Loss: 0.11164677143096924\n",
      "Epoch [25/200], Loss: 0.10296566784381866\n",
      "Epoch [26/200], Loss: 0.09518710523843765\n",
      "Epoch [27/200], Loss: 0.08982985466718674\n",
      "Epoch [28/200], Loss: 0.08321435749530792\n",
      "Epoch [29/200], Loss: 0.07733117789030075\n",
      "Epoch [30/200], Loss: 0.07184430211782455\n",
      "Epoch [31/200], Loss: 0.0663248598575592\n",
      "Epoch [32/200], Loss: 0.06211206316947937\n",
      "Epoch [33/200], Loss: 0.0579068660736084\n",
      "Epoch [34/200], Loss: 0.054606616497039795\n",
      "Epoch [35/200], Loss: 0.051191166043281555\n",
      "Epoch [36/200], Loss: 0.047771405428647995\n",
      "Epoch [37/200], Loss: 0.045270729809999466\n",
      "Epoch [38/200], Loss: 0.04388266056776047\n",
      "Epoch [39/200], Loss: 0.04167153686285019\n",
      "Epoch [40/200], Loss: 0.03903637081384659\n",
      "Epoch [41/200], Loss: 0.03716022148728371\n",
      "Epoch [42/200], Loss: 0.035668425261974335\n",
      "Epoch [43/200], Loss: 0.03426579385995865\n",
      "Epoch [44/200], Loss: 0.0328022837638855\n",
      "Epoch [45/200], Loss: 0.03179449960589409\n",
      "Epoch [46/200], Loss: 0.030734049156308174\n",
      "Epoch [47/200], Loss: 0.029527556151151657\n",
      "Epoch [48/200], Loss: 0.028938110917806625\n",
      "Epoch [49/200], Loss: 0.027805780991911888\n",
      "Epoch [50/200], Loss: 0.02714334987103939\n",
      "Epoch [51/200], Loss: 0.026013001799583435\n",
      "Epoch [52/200], Loss: 0.026148412376642227\n",
      "Epoch [53/200], Loss: 0.02501417137682438\n",
      "Epoch [54/200], Loss: 0.024782532826066017\n",
      "Epoch [55/200], Loss: 0.023976940661668777\n",
      "Epoch [56/200], Loss: 0.02344522438943386\n",
      "Epoch [57/200], Loss: 0.02298690192401409\n",
      "Epoch [58/200], Loss: 0.022800864651799202\n",
      "Epoch [59/200], Loss: 0.02203877829015255\n",
      "Epoch [60/200], Loss: 0.02186082862317562\n",
      "Epoch [61/200], Loss: 0.02145547792315483\n",
      "Epoch [62/200], Loss: 0.02111400105059147\n",
      "Epoch [63/200], Loss: 0.021111004054546356\n",
      "Epoch [64/200], Loss: 0.021235451102256775\n",
      "Epoch [65/200], Loss: 0.020454218611121178\n",
      "Epoch [66/200], Loss: 0.020284976810216904\n",
      "Epoch [67/200], Loss: 0.019960373640060425\n",
      "Epoch [68/200], Loss: 0.019969569519162178\n",
      "Epoch [69/200], Loss: 0.019146228209137917\n",
      "Epoch [70/200], Loss: 0.01951329968869686\n",
      "Epoch [71/200], Loss: 0.01905273087322712\n",
      "Epoch [72/200], Loss: 0.018939871340990067\n",
      "Epoch [73/200], Loss: 0.018563639372587204\n",
      "Epoch [74/200], Loss: 0.018615636974573135\n",
      "Epoch [75/200], Loss: 0.018304886296391487\n",
      "Epoch [76/200], Loss: 0.01797996088862419\n",
      "Epoch [77/200], Loss: 0.017887452617287636\n",
      "Epoch [78/200], Loss: 0.017940785735845566\n",
      "Epoch [79/200], Loss: 0.017486266791820526\n",
      "Epoch [80/200], Loss: 0.01746840588748455\n",
      "Epoch [81/200], Loss: 0.017102232202887535\n",
      "Epoch [82/200], Loss: 0.017105089500546455\n",
      "Epoch [83/200], Loss: 0.016733380034565926\n",
      "Epoch [84/200], Loss: 0.016671737655997276\n",
      "Epoch [85/200], Loss: 0.016277296468615532\n",
      "Epoch [86/200], Loss: 0.01615353487432003\n",
      "Epoch [87/200], Loss: 0.016070151701569557\n",
      "Epoch [88/200], Loss: 0.015822922810912132\n",
      "Epoch [89/200], Loss: 0.015388056635856628\n",
      "Epoch [90/200], Loss: 0.015657253563404083\n",
      "Epoch [91/200], Loss: 0.015209450386464596\n",
      "Epoch [92/200], Loss: 0.01531938835978508\n",
      "Epoch [93/200], Loss: 0.015104217454791069\n",
      "Epoch [94/200], Loss: 0.015498699620366096\n",
      "Epoch [95/200], Loss: 0.01536652259528637\n",
      "Epoch [96/200], Loss: 0.015862323343753815\n",
      "Epoch [97/200], Loss: 0.015583674423396587\n",
      "Epoch [98/200], Loss: 0.014587637968361378\n",
      "Epoch [99/200], Loss: 0.014702104032039642\n",
      "Epoch [100/200], Loss: 0.014520714059472084\n",
      "Epoch [101/200], Loss: 0.014230665750801563\n",
      "Epoch [102/200], Loss: 0.014032993465662003\n",
      "Epoch [103/200], Loss: 0.014286665245890617\n",
      "Epoch [104/200], Loss: 0.01409598346799612\n",
      "Epoch [105/200], Loss: 0.013588094152510166\n",
      "Epoch [106/200], Loss: 0.013453097082674503\n",
      "Epoch [107/200], Loss: 0.013409607112407684\n",
      "Epoch [108/200], Loss: 0.013149403035640717\n",
      "Epoch [109/200], Loss: 0.013141301460564137\n",
      "Epoch [110/200], Loss: 0.01283340621739626\n",
      "Epoch [111/200], Loss: 0.012824492529034615\n",
      "Epoch [112/200], Loss: 0.012342739850282669\n",
      "Epoch [113/200], Loss: 0.012487290427088737\n",
      "Epoch [114/200], Loss: 0.012282567098736763\n",
      "Epoch [115/200], Loss: 0.012123352847993374\n",
      "Epoch [116/200], Loss: 0.012309921905398369\n",
      "Epoch [117/200], Loss: 0.011849000118672848\n",
      "Epoch [118/200], Loss: 0.011969990096986294\n",
      "Epoch [119/200], Loss: 0.011744375340640545\n",
      "Epoch [120/200], Loss: 0.011610714718699455\n",
      "Epoch [121/200], Loss: 0.011537344194948673\n",
      "Epoch [122/200], Loss: 0.011500836350023746\n",
      "Epoch [123/200], Loss: 0.011424414813518524\n",
      "Epoch [124/200], Loss: 0.011428999714553356\n",
      "Epoch [125/200], Loss: 0.011216061189770699\n",
      "Epoch [126/200], Loss: 0.011091525666415691\n",
      "Epoch [127/200], Loss: 0.011024504899978638\n",
      "Epoch [128/200], Loss: 0.011121418327093124\n",
      "Epoch [129/200], Loss: 0.010962490923702717\n",
      "Epoch [130/200], Loss: 0.010953482240438461\n",
      "Epoch [131/200], Loss: 0.01096632331609726\n",
      "Epoch [132/200], Loss: 0.01100214198231697\n",
      "Epoch [133/200], Loss: 0.010964350774884224\n",
      "Epoch [134/200], Loss: 0.010790727101266384\n",
      "Epoch [135/200], Loss: 0.010693101212382317\n",
      "Epoch [136/200], Loss: 0.010713927447795868\n",
      "Epoch [137/200], Loss: 0.010633900761604309\n",
      "Epoch [138/200], Loss: 0.010555696673691273\n",
      "Epoch [139/200], Loss: 0.010487300343811512\n",
      "Epoch [140/200], Loss: 0.010380434803664684\n",
      "Epoch [141/200], Loss: 0.010355514474213123\n",
      "Epoch [142/200], Loss: 0.010410982184112072\n",
      "Epoch [143/200], Loss: 0.010351761244237423\n",
      "Epoch [144/200], Loss: 0.010418400168418884\n",
      "Epoch [145/200], Loss: 0.010332201607525349\n",
      "Epoch [146/200], Loss: 0.010343995876610279\n",
      "Epoch [147/200], Loss: 0.010427448898553848\n",
      "Epoch [148/200], Loss: 0.010323477908968925\n",
      "Epoch [149/200], Loss: 0.010058051906526089\n",
      "Epoch [150/200], Loss: 0.010284533724188805\n",
      "Epoch [151/200], Loss: 0.010071534663438797\n",
      "Epoch [152/200], Loss: 0.009985437616705894\n",
      "Epoch [153/200], Loss: 0.010022484697401524\n",
      "Epoch [154/200], Loss: 0.009919346310198307\n",
      "Epoch [155/200], Loss: 0.00982485618442297\n",
      "Epoch [156/200], Loss: 0.009892821311950684\n",
      "Epoch [157/200], Loss: 0.009983483701944351\n",
      "Epoch [158/200], Loss: 0.009957253001630306\n",
      "Epoch [159/200], Loss: 0.009947950020432472\n",
      "Epoch [160/200], Loss: 0.0098407082259655\n",
      "Epoch [161/200], Loss: 0.009712287224829197\n",
      "Epoch [162/200], Loss: 0.009856811724603176\n",
      "Epoch [163/200], Loss: 0.009634727612137794\n",
      "Epoch [164/200], Loss: 0.009648673236370087\n",
      "Epoch [165/200], Loss: 0.009548823349177837\n",
      "Epoch [166/200], Loss: 0.009575641714036465\n",
      "Epoch [167/200], Loss: 0.009495118632912636\n",
      "Epoch [168/200], Loss: 0.009415577165782452\n",
      "Epoch [169/200], Loss: 0.009466130286455154\n",
      "Epoch [170/200], Loss: 0.009326552972197533\n",
      "Epoch [171/200], Loss: 0.009429184719920158\n",
      "Epoch [172/200], Loss: 0.009482380002737045\n",
      "Epoch [173/200], Loss: 0.009315679781138897\n",
      "Epoch [174/200], Loss: 0.00941756833344698\n",
      "Epoch [175/200], Loss: 0.009411154314875603\n",
      "Epoch [176/200], Loss: 0.009331178851425648\n",
      "Epoch [177/200], Loss: 0.009307712316513062\n",
      "Epoch [178/200], Loss: 0.009414409287273884\n",
      "Epoch [179/200], Loss: 0.009246723726391792\n",
      "Epoch [180/200], Loss: 0.009288187138736248\n",
      "Epoch [181/200], Loss: 0.009330397471785545\n",
      "Epoch [182/200], Loss: 0.009157110005617142\n",
      "Epoch [183/200], Loss: 0.009136127308011055\n",
      "Epoch [184/200], Loss: 0.009142176248133183\n",
      "Epoch [185/200], Loss: 0.009172348305583\n",
      "Epoch [186/200], Loss: 0.009076423943042755\n",
      "Epoch [187/200], Loss: 0.009237755089998245\n",
      "Epoch [188/200], Loss: 0.008973689749836922\n",
      "Epoch [189/200], Loss: 0.009201562032103539\n",
      "Epoch [190/200], Loss: 0.008966690860688686\n",
      "Epoch [191/200], Loss: 0.009020267985761166\n",
      "Epoch [192/200], Loss: 0.008946004323661327\n",
      "Epoch [193/200], Loss: 0.00903538428246975\n",
      "Epoch [194/200], Loss: 0.00894266925752163\n",
      "Epoch [195/200], Loss: 0.008976456709206104\n",
      "Epoch [196/200], Loss: 0.008941787295043468\n",
      "Epoch [197/200], Loss: 0.00887981429696083\n",
      "Epoch [198/200], Loss: 0.008928196504712105\n",
      "Epoch [199/200], Loss: 0.008890601806342602\n",
      "Epoch [200/200], Loss: 0.008821516297757626\n"
     ]
    }
   ],
   "source": [
    "# Initialize model\n",
    "input_size = 512\n",
    "output_size = 512\n",
    "n_layers = 2\n",
    "n_heads = 4\n",
    "hidden_size = 32\n",
    "dropout = 0.1\n",
    "model = TransformerModel(input_size, output_size, n_layers, n_heads, hidden_size, dropout).to(device)\n",
    "\n",
    "\n",
    "# Convert data to the appropriate device\n",
    "X_train_tensor = X_train_tensor.to(device)\n",
    "y_train_tensor = y_train_tensor.to(device)\n",
    "X_test_tensor = X_test_tensor.to(device)\n",
    "y_test_tensor = y_test_tensor.to(device)\n",
    "\n",
    "\n",
    "# Define loss function and optimizer\n",
    "criterion = nn.MSELoss()\n",
    "optimizer = optim.Adam(model.parameters(), lr=0.001)\n",
    "\n",
    "# Train the model\n",
    "n_epochs = 200\n",
    "for epoch in range(n_epochs):\n",
    "    optimizer.zero_grad()\n",
    "    y_pred = model(X_train_tensor)\n",
    "    loss = criterion(y_pred, y_train_tensor)\n",
    "    loss.backward()\n",
    "    optimizer.step()\n",
    "    print(f\"Epoch [{epoch+1}/{n_epochs}], Loss: {loss.item()}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test MSE: 0.009519089013338089\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the model\n",
    "with torch.no_grad():\n",
    "    model.eval()\n",
    "    y_pred_test = model(X_test_tensor)\n",
    "    y_pred_test = y_pred_test.cpu()\n",
    "    mse = mean_squared_error(y_test, y_pred_test)\n",
    "    print(f\"Test MSE: {mse}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot predictions\n",
    "plt.figure(figsize=(10, 6))\n",
    "plt.plot(y_test[0], label=\"True\")\n",
    "plt.plot(y_pred_test[0], label=\"Predicted\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Comment\n",
    "\n",
    "Result over-fit, but it works."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "torch210",
   "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": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}