day02/Seminar_RNN-LSTM.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> Code from https://machinelearningmastery.com/lstm-for-time-series-prediction-in-pytorch/ with modifications.\n",
    "\n",
    "In this seminar, we work with the \"airline passengers\" dataset, indicating the number of passengers over time. Let's start by loading the data!\n",
    "\n",
    "Link: https://raw.githubusercontent.com/jbrownlee/Datasets/master/airline-passengers.csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Month</th>\n",
       "      <th>Passengers</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1949-01</td>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1949-02</td>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1949-03</td>\n",
       "      <td>132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1949-04</td>\n",
       "      <td>129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1949-05</td>\n",
       "      <td>121</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Month  Passengers\n",
       "0  1949-01         112\n",
       "1  1949-02         118\n",
       "2  1949-03         132\n",
       "3  1949-04         129\n",
       "4  1949-05         121"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "df = pd.read_csv('airline-passengers.csv')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[112.],\n",
       "       [118.],\n",
       "       [132.],\n",
       "       [129.],\n",
       "       [121.],\n",
       "       [135.],\n",
       "       [148.],\n",
       "       [148.],\n",
       "       [136.],\n",
       "       [119.],\n",
       "       [104.],\n",
       "       [118.],\n",
       "       [115.],\n",
       "       [126.],\n",
       "       [141.],\n",
       "       [135.],\n",
       "       [125.],\n",
       "       [149.],\n",
       "       [170.],\n",
       "       [170.],\n",
       "       [158.],\n",
       "       [133.],\n",
       "       [114.],\n",
       "       [140.],\n",
       "       [145.],\n",
       "       [150.],\n",
       "       [178.],\n",
       "       [163.],\n",
       "       [172.],\n",
       "       [178.],\n",
       "       [199.],\n",
       "       [199.],\n",
       "       [184.],\n",
       "       [162.],\n",
       "       [146.],\n",
       "       [166.],\n",
       "       [171.],\n",
       "       [180.],\n",
       "       [193.],\n",
       "       [181.],\n",
       "       [183.],\n",
       "       [218.],\n",
       "       [230.],\n",
       "       [242.],\n",
       "       [209.],\n",
       "       [191.],\n",
       "       [172.],\n",
       "       [194.],\n",
       "       [196.],\n",
       "       [196.],\n",
       "       [236.],\n",
       "       [235.],\n",
       "       [229.],\n",
       "       [243.],\n",
       "       [264.],\n",
       "       [272.],\n",
       "       [237.],\n",
       "       [211.],\n",
       "       [180.],\n",
       "       [201.],\n",
       "       [204.],\n",
       "       [188.],\n",
       "       [235.],\n",
       "       [227.],\n",
       "       [234.],\n",
       "       [264.],\n",
       "       [302.],\n",
       "       [293.],\n",
       "       [259.],\n",
       "       [229.],\n",
       "       [203.],\n",
       "       [229.],\n",
       "       [242.],\n",
       "       [233.],\n",
       "       [267.],\n",
       "       [269.],\n",
       "       [270.],\n",
       "       [315.],\n",
       "       [364.],\n",
       "       [347.],\n",
       "       [312.],\n",
       "       [274.],\n",
       "       [237.],\n",
       "       [278.],\n",
       "       [284.],\n",
       "       [277.],\n",
       "       [317.],\n",
       "       [313.],\n",
       "       [318.],\n",
       "       [374.],\n",
       "       [413.],\n",
       "       [405.],\n",
       "       [355.],\n",
       "       [306.],\n",
       "       [271.],\n",
       "       [306.],\n",
       "       [315.],\n",
       "       [301.],\n",
       "       [356.],\n",
       "       [348.],\n",
       "       [355.],\n",
       "       [422.],\n",
       "       [465.],\n",
       "       [467.],\n",
       "       [404.],\n",
       "       [347.],\n",
       "       [305.],\n",
       "       [336.],\n",
       "       [340.],\n",
       "       [318.],\n",
       "       [362.],\n",
       "       [348.],\n",
       "       [363.],\n",
       "       [435.],\n",
       "       [491.],\n",
       "       [505.],\n",
       "       [404.],\n",
       "       [359.],\n",
       "       [310.],\n",
       "       [337.],\n",
       "       [360.],\n",
       "       [342.],\n",
       "       [406.],\n",
       "       [396.],\n",
       "       [420.],\n",
       "       [472.],\n",
       "       [548.],\n",
       "       [559.],\n",
       "       [463.],\n",
       "       [407.],\n",
       "       [362.],\n",
       "       [405.],\n",
       "       [417.],\n",
       "       [391.],\n",
       "       [419.],\n",
       "       [461.],\n",
       "       [472.],\n",
       "       [535.],\n",
       "       [622.],\n",
       "       [606.],\n",
       "       [508.],\n",
       "       [461.],\n",
       "       [390.],\n",
       "       [432.]], dtype=float32)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_points = df[[\"Passengers\"]].values.astype('float32')\n",
    "data_points"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "plt.plot(data_points)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/15/bmbcw71s1l12pc4g0617fwww0000gn/T/ipykernel_40783/826358318.py:13: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at  /Users/runner/work/pytorch/pytorch/pytorch/torch/csrc/utils/tensor_new.cpp:204.)\n",
      "  X_train = torch.tensor(X_train)\n"
     ]
    }
   ],
   "source": [
    "train_size = int(len(data_points) * 0.8)\n",
    "train_data_points = data_points[:train_size]\n",
    "test_data_points = data_points[train_size:]\n",
    "\n",
    "sequence_size = 4\n",
    "\n",
    "import torch\n",
    "\n",
    "X_train, y_train = [], []\n",
    "for i in range(len(train_data_points) - sequence_size):\n",
    "    X_train.append(train_data_points[i:i+sequence_size])\n",
    "    y_train.append(train_data_points[i+1:i+sequence_size+1])\n",
    "X_train = torch.tensor(X_train)\n",
    "y_train = torch.tensor(y_train)\n",
    "\n",
    "X_test, y_test = [], []\n",
    "for i in range(len(test_data_points) - sequence_size):\n",
    "    X_test.append(test_data_points[i:i+sequence_size])\n",
    "    y_test.append(test_data_points[i+1:i+sequence_size+1])\n",
    "X_test = torch.tensor(X_test)\n",
    "y_test = torch.tensor(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.nn as nn\n",
    "from tqdm import tqdm\n",
    "\n",
    "class SimpleRNNModel(nn.Module):\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        self.rnn = nn.RNN(input_size=1, hidden_size=64, num_layers=1, batch_first=True) # batch_first=True: (batch, seq_length, features), the format that comes from DataLoader\n",
    "        self.linear = nn.Linear(64, 1)\n",
    "    def forward(self, x):\n",
    "        x, _ = self.rnn(x)\n",
    "        x = self.linear(x)\n",
    "        return x"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As https://towardsdatascience.com/pytorch-basics-how-to-train-your-neural-net-intro-to-rnn-cb6ebc594677 says regarding the outputs of the `rnn` function:\n",
    "\n",
    "\n",
    "- out is the output of the RNN from all timesteps from the last RNN layer. It is of the size (seq_len, batch, num_directions * hidden_size). If batch_first=True, the output size is (batch, seq_len, num_directions * hidden_size).\n",
    "\n",
    "- h_n is the hidden value from the last time-step of all RNN layers. It is of the size (num_layers * num_directions, batch, hidden_size). h_n is unaffected by batch_first=True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "\n",
    "def training_loop(model, number_of_epochs = 1000):\n",
    "    loader = torch.utils.data.DataLoader(torch.utils.data.TensorDataset(X_train, y_train), shuffle=True)\n",
    "    loss_function = nn.MSELoss() # from PyTorch documentation: measures the mean squared error (squared L2 norm) between each element in the input x and target y\n",
    "    optimizer = torch.optim.Adam(model.parameters())\n",
    "    for epoch in tqdm(range(number_of_epochs)):\n",
    "        model.train()\n",
    "        for X_batch, y_batch in loader:\n",
    "            y_pred = model(X_batch)\n",
    "            loss = loss_function(y_pred, y_batch)\n",
    "            optimizer.zero_grad() # for every mini batch in training, set gradients to zero before the backprop\n",
    "            loss.backward()\n",
    "            optimizer.step()\n",
    "        # Validation\n",
    "        if epoch % 100 == 0 or epoch == number_of_epochs - 1:\n",
    "            model.eval()\n",
    "            with torch.no_grad(): # do not update the gradients -> no effect on backprop\n",
    "                y_pred_train = model(X_train)\n",
    "                y_pred_test = model(X_test)\n",
    "                print(f\"RMSE for epoch {epoch}: train {np.sqrt(loss_function(y_pred_train, y_train))}, test {np.sqrt(loss_function(y_pred_test, y_test))}\")\n",
    "\n",
    "    with torch.no_grad():\n",
    "        # shift train predictions for plotting\n",
    "        train_results = np.ones_like(data_points) * np.nan # an array with the same shape as data_points, filled with NaNs. Stores the training predictions for plotting (if NaN, it is not plotted)\n",
    "        train_results[4:train_size] = model(X_train)[:, -1, :] # only keep the last prediction for each sequence in the batch\n",
    "        # shift test predictions for plotting\n",
    "        test_results = np.ones_like(data_points) * np.nan\n",
    "        test_results[train_size+4:len(data_points)] = model(X_test)[:, -1, :]\n",
    "    # plot\n",
    "    plt.plot(data_points, c='r')\n",
    "    plt.plot(train_results, c='g')\n",
    "    plt.plot(test_results, c='b')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From https://stackoverflow.com/questions/51433378/what-does-model-train-do-in-pytorch:\n",
    "\n",
    "model.train() tells your model that you are training the model. This helps inform layers such as Dropout and BatchNorm, which are designed to behave differently during training and evaluation. For instance, in training mode, BatchNorm updates a moving average on each new batch; whereas, for evaluation mode, these updates are frozen."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|          | 4/1000 [00:00<00:56, 17.53it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 0: train 248.2376251220703, test 442.794189453125\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 11%|█         | 106/1000 [00:04<00:31, 28.67it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 100: train 36.80834197998047, test 144.93267822265625\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 20%|██        | 205/1000 [00:08<00:28, 28.01it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 200: train 29.197769165039062, test 109.51982116699219\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 30%|███       | 303/1000 [00:13<00:47, 14.79it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 300: train 27.806264877319336, test 94.8164291381836\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 41%|████      | 406/1000 [00:17<00:20, 29.18it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 400: train 27.8165283203125, test 99.3706283569336\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 50%|█████     | 505/1000 [00:20<00:17, 27.89it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 500: train 25.77418327331543, test 90.3497543334961\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 60%|██████    | 604/1000 [00:24<00:13, 29.14it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 600: train 27.775592803955078, test 86.20764923095703\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 71%|███████   | 706/1000 [00:27<00:10, 28.59it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 700: train 26.038999557495117, test 88.95542907714844\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 81%|████████  | 806/1000 [00:31<00:06, 29.53it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 800: train 27.822477340698242, test 94.7621841430664\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 90%|█████████ | 904/1000 [00:34<00:03, 26.85it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 900: train 31.177526473999023, test 87.74742889404297\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [00:38<00:00, 26.00it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 999: train 25.891618728637695, test 84.2864990234375\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = SimpleRNNModel()\n",
    "training_loop(model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|          | 3/1000 [00:00<01:17, 12.90it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 0: train 251.66552734375, test 446.4869384765625\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 10%|█         | 103/1000 [00:07<00:59, 15.20it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 100: train 41.86274337768555, test 152.73178100585938\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 20%|██        | 203/1000 [00:14<00:58, 13.57it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 200: train 27.32361602783203, test 91.3949203491211\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 30%|███       | 303/1000 [00:21<00:45, 15.21it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 300: train 25.076194763183594, test 85.40010070800781\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 40%|████      | 403/1000 [00:28<00:39, 15.09it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 400: train 27.000324249267578, test 88.91984558105469\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 50%|█████     | 503/1000 [00:35<00:34, 14.54it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 500: train 27.032108306884766, test 84.3886947631836\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 60%|██████    | 603/1000 [00:42<00:28, 14.01it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 600: train 23.915729522705078, test 78.1270523071289\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 70%|███████   | 703/1000 [00:48<00:19, 15.21it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 700: train 23.228641510009766, test 72.24260711669922\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 80%|████████  | 803/1000 [00:55<00:13, 15.14it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 800: train 23.47812843322754, test 72.56974792480469\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 90%|█████████ | 903/1000 [01:02<00:06, 15.14it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 900: train 23.517175674438477, test 69.63007354736328\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [01:08<00:00, 14.58it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 999: train 22.876998901367188, test 71.02112579345703\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "class LSTMModel(nn.Module):\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        self.lstm = nn.LSTM(input_size=1, hidden_size=64, num_layers=1, batch_first=True)\n",
    "        self.linear = nn.Linear(64, 1)\n",
    "    def forward(self, x):\n",
    "        x, _ = self.lstm(x)\n",
    "        x = self.linear(x)\n",
    "        return x\n",
    "\n",
    "model = LSTMModel()\n",
    "training_loop(model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|          | 3/1000 [00:00<01:38, 10.13it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 0: train 248.76239013671875, test 443.3445129394531\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 10%|█         | 103/1000 [00:06<00:51, 17.42it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 100: train 36.96455001831055, test 137.48960876464844\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 20%|██        | 203/1000 [00:11<00:46, 17.23it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 200: train 26.418964385986328, test 90.94501495361328\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 30%|███       | 303/1000 [00:17<00:39, 17.84it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 300: train 25.757009506225586, test 88.41075134277344\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 40%|████      | 403/1000 [00:23<00:33, 17.72it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 400: train 26.63245391845703, test 85.4991683959961\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 50%|█████     | 503/1000 [00:29<00:27, 17.83it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 500: train 26.07852554321289, test 87.52980041503906\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 60%|██████    | 603/1000 [00:35<00:22, 17.74it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 600: train 27.62359619140625, test 88.74533081054688\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 70%|███████   | 703/1000 [00:40<00:16, 17.78it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 700: train 25.480579376220703, test 86.0707778930664\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 80%|████████  | 803/1000 [00:46<00:11, 17.74it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 800: train 25.72702407836914, test 78.04386138916016\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 90%|█████████ | 903/1000 [00:52<00:05, 17.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 900: train 28.141357421875, test 82.66927337646484\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [00:59<00:00, 16.83it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for epoch 999: train 26.156238555908203, test 75.82245635986328\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "class GRUModel(nn.Module):\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        self.gru = nn.GRU(input_size=1, hidden_size=64, num_layers=1, batch_first=True)\n",
    "        self.linear = nn.Linear(64, 1)\n",
    "    def forward(self, x):\n",
    "        x, _ = self.gru(x)\n",
    "        x = self.linear(x)\n",
    "        return x\n",
    "    \n",
    "model = GRUModel()\n",
    "training_loop(model)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "env_tf",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}