eth_pred_baseline.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "frequent-motorcycle",
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "import os\n",
    "import csv\n",
    "import urllib\n",
    "from io import BytesIO\n",
    "from PIL import Image\n",
    "\n",
    "from socket import timeout\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import torch\n",
    "from torchvision import models\n",
    "from torch.utils.data import Dataset, SubsetRandomSampler\n",
    "from torchvision import transforms\n",
    "import pickle\n",
    "from plot_tool import *\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "import seaborn as sns\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "valued-louisville",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The autoreload extension is already loaded. To reload it, use:\n",
      "  %reload_ext autoreload\n"
     ]
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "id": "shared-beast",
   "metadata": {},
   "outputs": [],
   "source": [
    "device = torch.device(\"cuda:0\") #Подрубайтунг"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "id": "corporate-outline",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_folder = \"stats/\"\n",
    "points= 20"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "id": "alien-contrary",
   "metadata": {},
   "outputs": [],
   "source": [
    "def saveStats(data, name):\n",
    "    with open('stats/'+str(name)+\".stat\", 'wb') as f:\n",
    "        pickle.dump(data, f)\n",
    "        \n",
    "def readStats(name):\n",
    "    with open('stats/'+str(name), 'rb') as f:\n",
    "        return pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "id": "phantom-botswana",
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = MinMaxScaler(feature_range=(-1, 1))\n",
    "\n",
    "class CryptoDataset(Dataset):\n",
    "    def __init__(self, folder, transform=None, cutoff = 25):\n",
    "        self.transform = transform\n",
    "        self.folder = folder\n",
    "        \n",
    "        files = os.listdir(self.folder)\n",
    "        \n",
    "        self.count = len(files)\n",
    "        self.games_path = list()\n",
    "        self.cutoff = cutoff\n",
    "        \n",
    "        for f in os.listdir(self.folder):\n",
    "            self.games_path.append(f)\n",
    "\n",
    "    def __len__(self):\n",
    "        return self.count\n",
    "\n",
    "    def __getitem__(self, index):\n",
    "        path = self.games_path[index]\n",
    "        \n",
    "        base_stats = readStats(path)\n",
    "\n",
    "        close_train = np.array(base_stats.closes.getData()[:self.cutoff])\n",
    "        close_gt = np.array(base_stats.closes.getData()[self.cutoff:])\n",
    "        \n",
    "        trainOut = np.array([close_train, close_train])\n",
    "        gtOut = np.array([close_gt, close_gt])\n",
    "        print(trainOut.shape)\n",
    "        trainOut = torch.from_numpy(trainOut).type(torch.Tensor)\n",
    "        gtOut = torch.from_numpy(gtOut).type(torch.Tensor)\n",
    "           \n",
    "        return trainOut, gtOut, index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "id": "offshore-jamaica",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dataset = CryptoDataset(train_folder, transform=transforms.Compose([transforms.ToTensor(),]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "id": "civil-prime",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10"
      ]
     },
     "execution_count": 201,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train_dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "id": "lovely-traffic",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2, 25)\n",
      "torch.Size([2, 25]) torch.Size([2, 5])\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x2cd819f4cd0>]"
      ]
     },
     "execution_count": 202,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"whitegrid\")\n",
    "a, b, c = train_dataset[0]\n",
    "print(a.shape, b.shape)\n",
    "res = np.concatenate([a[0], b[0]])\n",
    "plt.plot(res)\n",
    "plt.plot(a[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "id": "directed-assembly",
   "metadata": {},
   "outputs": [],
   "source": [
    "class GRU(nn.Module):\n",
    "    def __init__(self, input_dim, hidden_dim, num_layers, output_dim):\n",
    "        super(GRU, self).__init__()\n",
    "        self.hidden_dim = hidden_dim\n",
    "        self.num_layers = num_layers\n",
    "        \n",
    "        self.gru = nn.GRU(input_dim, hidden_dim, num_layers, batch_first=True)\n",
    "        self.fc = nn.Linear(hidden_dim, output_dim)\n",
    "\n",
    "    def forward(self, x):\n",
    "        h0 = torch.zeros(self.num_layers, x.size(0), self.hidden_dim).requires_grad_()\n",
    "        out, (hn) = self.gru(x, (h0.detach()))\n",
    "        out = self.fc(out[:, -1, :]) \n",
    "        return out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "id": "boolean-hostel",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10\n"
     ]
    }
   ],
   "source": [
    "batch_size = 1\n",
    "\n",
    "data_size = len(train_dataset)\n",
    "print(data_size)\n",
    "validation_fraction = .2\n",
    "\n",
    "\n",
    "val_split = int(np.floor((validation_fraction) * data_size))\n",
    "indices = list(range(data_size))\n",
    "np.random.seed(42)\n",
    "np.random.shuffle(indices)\n",
    "\n",
    "val_indices, train_indices = indices[:val_split], indices[val_split:]\n",
    "\n",
    "train_sampler = SubsetRandomSampler(train_indices)\n",
    "val_sampler = SubsetRandomSampler(val_indices)\n",
    "\n",
    "train_loader = torch.utils.data.DataLoader(train_dataset, batch_size=batch_size, \n",
    "                                           sampler=train_sampler)\n",
    "val_loader = torch.utils.data.DataLoader(train_dataset, batch_size=batch_size,\n",
    "                                         sampler=val_sampler)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "id": "existing-newport",
   "metadata": {},
   "outputs": [],
   "source": [
    "#for i_step, (x, y, z) in enumerate(train_loader):\n",
    "#    print(i_step, x, y, z)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "id": "engaging-bunny",
   "metadata": {},
   "outputs": [],
   "source": [
    "import copy\n",
    "\n",
    "def compute_accuracy(model, loader, device):\n",
    "    \"\"\"\n",
    "    Computes accuracy on the dataset wrapped in a loader\n",
    "\n",
    "    Returns: accuracy as a float value between 0 and 1\n",
    "    \"\"\"\n",
    "    model.eval()  # Evaluation mode\n",
    "\n",
    "    correct = 0\n",
    "    total = 0\n",
    "    with torch.no_grad():\n",
    "        for i_step, (x, y, _) in enumerate(loader):\n",
    "            x, y = x.float(), y.float()\n",
    "            prediction = model(x.to(device)).cpu()\n",
    "            #print(prediction.shape)\n",
    "            \n",
    "            \n",
    "    \n",
    "    return 1.0\n",
    "\n",
    "def train_model_v2(model, train_loader, val_loader, loss, optimizer, num_epochs, scheduler = None):    \n",
    "    loss_history = []\n",
    "    train_history = []\n",
    "    val_history = []\n",
    "    for epoch in range(num_epochs):\n",
    "        model.train() # Enter train mode\n",
    "        \n",
    "        loss_accum = 0\n",
    "        correct_samples = 0\n",
    "        total_samples = 0\n",
    "        for i_step, (x, y,_) in enumerate(train_loader):\n",
    "            x=x.float()\n",
    "            y=y.float()\n",
    "            print(x)\n",
    "            x_gpu = x.to(device)\n",
    "            y_gpu = y.to(device)\n",
    "            \n",
    "            prediction = model(x_gpu)    \n",
    "            loss_value = loss(prediction, y_gpu)\n",
    "            \n",
    "            optimizer.zero_grad()\n",
    "            loss_value.backward()\n",
    "            optimizer.step()\n",
    "\n",
    "            for i in range(x.shape[0]):\n",
    "                for j in range(x.shape[1]):\n",
    "                    y_data = y[i][j]\n",
    "                    pred_data = np.around(prediction.cpu().detach().numpy()[i][j])\n",
    "                    countMap = map(lambda x, y: 1 if x==y else 0, y_data.reshape(-1), pred_data.reshape(-1))\n",
    "                    #print(pred_data)\n",
    "\n",
    "                    correct_samples += sum(list(countMap))/(y_data.shape[0]*y_data.shape[1])\n",
    "                    total_samples += 1\n",
    "            \n",
    "            loss_accum += loss_value\n",
    "            \n",
    "        ave_loss = loss_accum / i_step\n",
    "        \n",
    "        train_accuracy = float(correct_samples) / total_samples\n",
    "        val_accuracy = compute_accuracy(model, val_loader, device)\n",
    "        \n",
    "        if scheduler is not None:\n",
    "            scheduler.step(ave_loss)\n",
    "            \n",
    "        loss_history.append(float(ave_loss))\n",
    "        train_history.append(train_accuracy)\n",
    "        val_history.append(val_accuracy)\n",
    "        \n",
    "        print(\"Average loss: %f, Train accuracy: %f, Val accuracy: %f\" % (ave_loss, train_accuracy, val_accuracy))\n",
    "        \n",
    "    return loss_history, train_history, val_history"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "id": "filled-lender",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "GRU(\n",
      "  (gru): GRU(1, 32, num_layers=2, batch_first=True)\n",
      "  (fc): Linear(in_features=32, out_features=1, bias=True)\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "model = GRU(input_dim=1, hidden_dim=32, output_dim=1, num_layers=2)\n",
    "print(model)\n",
    "criterion = torch.nn.MSELoss(reduction='mean')\n",
    "optimiser = torch.optim.Adam(model.parameters(), lr=0.01)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "id": "spare-monitor",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2, 25)\n",
      "tensor([[[2709.4163, 2687.2722, 2766.0435, 2801.0090, 2844.2583, 2819.3560,\n",
      "          2806.4170, 2883.1008, 2897.9602, 2847.5176, 2850.9915, 2830.7952,\n",
      "          2840.8364, 2860.6201, 2712.2827, 2757.6536, 2749.4617, 2713.6199,\n",
      "          2748.1006, 2760.3040, 2751.0054, 2804.4956, 2821.3608, 2849.0737,\n",
      "          2892.3008],\n",
      "         [2709.4163, 2687.2722, 2766.0435, 2801.0090, 2844.2583, 2819.3560,\n",
      "          2806.4170, 2883.1008, 2897.9602, 2847.5176, 2850.9915, 2830.7952,\n",
      "          2840.8364, 2860.6201, 2712.2827, 2757.6536, 2749.4617, 2713.6199,\n",
      "          2748.1006, 2760.3040, 2751.0054, 2804.4956, 2821.3608, 2849.0737,\n",
      "          2892.3008]]])\n"
     ]
    },
    {
     "ename": "RuntimeError",
     "evalue": "input.size(-1) must be equal to input_size. Expected 1, got 25",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-211-c14635ed7054>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mloss_history\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtrain_history\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mval_history\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtrain_model_v2\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtrain_loader\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mval_loader\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcriterion\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0moptimiser\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m50\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;32m<ipython-input-209-cd985bcfa0c2>\u001b[0m in \u001b[0;36mtrain_model_v2\u001b[1;34m(model, train_loader, val_loader, loss, optimizer, num_epochs, scheduler)\u001b[0m\n\u001b[0;32m     38\u001b[0m             \u001b[0my_gpu\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mto\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdevice\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     39\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 40\u001b[1;33m             \u001b[0mprediction\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx_gpu\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     41\u001b[0m             \u001b[0mloss_value\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mloss\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mprediction\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my_gpu\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     42\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mc:\\users\\neo\\appdata\\local\\programs\\python\\python38\\lib\\site-packages\\torch\\nn\\modules\\module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[1;34m(self, *input, **kwargs)\u001b[0m\n\u001b[0;32m    887\u001b[0m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    888\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 889\u001b[1;33m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    890\u001b[0m         for hook in itertools.chain(\n\u001b[0;32m    891\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-203-b8e756f79b3e>\u001b[0m in \u001b[0;36mforward\u001b[1;34m(self, x)\u001b[0m\n\u001b[0;32m     10\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     11\u001b[0m         \u001b[0mh0\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mzeros\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mnum_layers\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msize\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mhidden_dim\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mrequires_grad_\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 12\u001b[1;33m         \u001b[0mout\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0mhn\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mgru\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0mh0\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdetach\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     13\u001b[0m         \u001b[0mout\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfc\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mout\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m-\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m:\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     14\u001b[0m         \u001b[1;32mreturn\u001b[0m \u001b[0mout\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mc:\\users\\neo\\appdata\\local\\programs\\python\\python38\\lib\\site-packages\\torch\\nn\\modules\\module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[1;34m(self, *input, **kwargs)\u001b[0m\n\u001b[0;32m    887\u001b[0m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    888\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 889\u001b[1;33m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    890\u001b[0m         for hook in itertools.chain(\n\u001b[0;32m    891\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mc:\\users\\neo\\appdata\\local\\programs\\python\\python38\\lib\\site-packages\\torch\\nn\\modules\\rnn.py\u001b[0m in \u001b[0;36mforward\u001b[1;34m(self, input, hx)\u001b[0m\n\u001b[0;32m    817\u001b[0m             \u001b[0mhx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpermute_hidden\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mhx\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0msorted_indices\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    818\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 819\u001b[1;33m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcheck_forward_args\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mhx\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch_sizes\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    820\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mbatch_sizes\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    821\u001b[0m             result = _VF.gru(input, hx, self._flat_weights, self.bias, self.num_layers,\n",
      "\u001b[1;32mc:\\users\\neo\\appdata\\local\\programs\\python\\python38\\lib\\site-packages\\torch\\nn\\modules\\rnn.py\u001b[0m in \u001b[0;36mcheck_forward_args\u001b[1;34m(self, input, hidden, batch_sizes)\u001b[0m\n\u001b[0;32m    224\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    225\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mcheck_forward_args\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0minput\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mTensor\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mhidden\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mTensor\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch_sizes\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mOptional\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mTensor\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 226\u001b[1;33m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcheck_input\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch_sizes\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    227\u001b[0m         \u001b[0mexpected_hidden_size\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_expected_hidden_size\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch_sizes\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    228\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mc:\\users\\neo\\appdata\\local\\programs\\python\\python38\\lib\\site-packages\\torch\\nn\\modules\\rnn.py\u001b[0m in \u001b[0;36mcheck_input\u001b[1;34m(self, input, batch_sizes)\u001b[0m\n\u001b[0;32m    200\u001b[0m                     expected_input_dim, input.dim()))\n\u001b[0;32m    201\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0minput_size\u001b[0m \u001b[1;33m!=\u001b[0m \u001b[0minput\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msize\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m-\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 202\u001b[1;33m             raise RuntimeError(\n\u001b[0m\u001b[0;32m    203\u001b[0m                 'input.size(-1) must be equal to input_size. Expected {}, got {}'.format(\n\u001b[0;32m    204\u001b[0m                     self.input_size, input.size(-1)))\n",
      "\u001b[1;31mRuntimeError\u001b[0m: input.size(-1) must be equal to input_size. Expected 1, got 25"
     ]
    }
   ],
   "source": [
    "loss_history, train_history, val_history = train_model_v2(model, train_loader, val_loader, criterion, optimiser, 50)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "meaning-imagination",
   "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": "3.8.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}