Upload 10 files

Copy of andrea-dd-dataLte100.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2dda5dce8674842673a56eac34a77051908f407ead66085b1b84c4d372cba2c
+size 146828479

README.md

@@ -0,0 +1,2 @@
+# MLSP-Project
+MLSP E9-205 project submission of Vura T N M Srinadh (24923) and Islavath Ganesh (24976)

proj2.ipynb

@@ -0,0 +1,843 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "f49d031a",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import torch\n",
+    "from torch_geometric.data import Data\n",
+    "import torch_geometric.transforms as T\n",
+    "from torch_geometric.data import InMemoryDataset\n",
+    "from torch_geometric.loader import DataLoader\n",
+    "import numpy as np\n",
+    "import time"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "cd1e01d3",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "path =\"Copy of andrea-dd-dataLte100.npy\"\n",
+    "# path=\"/content/drive/MyDrive/Barc Datasets/andrea-juslin-dataGt100.npy\"\n",
+    "data = np.load(path, allow_pickle=True)\n",
+    "data = data[None][0]\n",
+    "dataXorig = []\n",
+    "dataY = []\n",
+    "dataPe = []\n",
+    "for x, y, z in zip(data['x'], data['y'], data['z']):\n",
+    "  if y > 3: continue # We only want 0 and 1 labels for this activity\n",
+    "  dataXorig.append(x)\n",
+    "  dataY.append(y)\n",
+    "  dataPe.append(z['pe'])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "4f8c43ad",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "featLen = 30 # we are restricting to first thirty nearest neighbours\n",
+    "dataAtoms = torch.tensor(np.array([x[0][:featLen+1] for x in dataXorig]), dtype=torch.float)\n",
+    "dataX = torch.tensor(np.array([x[1][:featLen+1] for x in dataXorig]).reshape(len(dataAtoms),featLen+1,1), dtype=torch.float)\n",
+    "# dataY = torch.nn.functional.one_hot(torch.tensor(dataY, dtype=torch.long),num_classes=4).float()\n",
+    "dataY=torch.tensor(dataY, dtype=torch.long)\n",
+    "#dataX = dataAtoms"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "70d2090f",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "(torch.Size([42699, 31, 4]),\n",
+       " torch.Size([42699, 31, 1]),\n",
+       " torch.Size([42699]),\n",
+       " torch.Size([42699, 31, 3]))"
+      ]
+     },
+     "execution_count": 4,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "dataXAll = torch.concat((dataX, dataAtoms), axis=2)\n",
+    "dataXAll.shape, dataX.shape, dataY.shape, dataAtoms.shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "832769fa",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "tensor([0, 0, 0,  ..., 1, 1, 1])"
+      ]
+     },
+     "execution_count": 5,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "dataY"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "0d430c0d",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "posData = []\n",
+    "for dt, y in zip(dataAtoms, dataY):\n",
+    "  data = Data(x= dt, pos=dt, y = y)#, pre_transform=T.RadiusGraph(r=4.0), transform=T.Distance())\n",
+    "  data.validate(raise_on_error=True)\n",
+    "  posData.append(data)\n",
+    "distData = []\n",
+    "for dt, x, y in zip(dataAtoms, dataX, dataY):\n",
+    "  data = Data(x= x, pos=dt, y = y)#, pre_transform=T.RadiusGraph(r=4.0), transform=T.Distance())\n",
+    "  data.validate(raise_on_error=True)\n",
+    "  distData.append(data)\n",
+    "allData = posData"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "ba1c05ab",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "dir(data)\n",
+    "data.edge_weight"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "47afb2ee",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import torch\n",
+    "from torch_geometric.data import Data\n",
+    "import torch_geometric.transforms as T\n",
+    "from torch_geometric.data import InMemoryDataset\n",
+    "from torch_geometric.loader import DataLoader\n",
+    "import numpy as np\n",
+    "import time"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "id": "80b6c9ac",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "device=\"cuda\" if torch.cuda.is_available() else \"cpu\""
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "id": "06225cca",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "class MyDataset(InMemoryDataset):\n",
+    "    def __init__(self, root, data_list, transform=None, pre_transform=None, pre_filter=None):\n",
+    "        self.data_list = data_list\n",
+    "        super().__init__(root, transform, pre_transform, pre_filter)\n",
+    "        self.data, self.slices = torch.load(self.processed_paths[0])\n",
+    "\n",
+    "    @property\n",
+    "    def raw_file_names(self):\n",
+    "        return ['mydata']\n",
+    "\n",
+    "    @property\n",
+    "    def processed_file_names(self):\n",
+    "        return ['data.pt']\n",
+    "\n",
+    "    def download(self):\n",
+    "        # Download to `self.raw_dir`.\n",
+    "        pass\n",
+    "\n",
+    "    def process(self):\n",
+    "        # Read data into huge `Data` list.\n",
+    "        data_list = self.data_list\n",
+    "\n",
+    "        if self.pre_filter is not None:\n",
+    "            data_list = [data for data in data_list if self.pre_filter(data)]\n",
+    "\n",
+    "        if self.pre_transform is not None:\n",
+    "            data_list = [self.pre_transform(data) for data in data_list]\n",
+    "\n",
+    "        data, slices = self.collate(data_list)\n",
+    "        torch.save((data, slices), self.processed_paths[0])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "id": "91654650",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Processing...\n",
+      "Done!\n",
+      "C:\\Users\\SrinadhVura\\AppData\\Local\\Temp\\ipykernel_18728\\1359331624.py:5: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.\n",
+      "  self.data, self.slices = torch.load(self.processed_paths[0])\n"
+     ]
+    }
+   ],
+   "source": [
+    "# !rm -rf ./data/processed/\n",
+    "import shutil\n",
+    "shutil.rmtree('./data/processed/', ignore_errors=True)\n",
+    "dataset = MyDataset(\"./data\",allData, pre_transform=T.Compose([T.RadiusGraph(r=2.0), T.Distance()]))\n",
+    "#dataset = MyDataset(\"./data\",allData, pre_transform=T.Compose([T.RadiusGraph(r=2.0), T.Distance()]))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "id": "3edad4e2",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from torch_geometric.nn import GATv2Conv\n",
+    "import torch.nn.functional as F\n",
+    "from torch_geometric.nn.glob import global_mean_pool\n",
+    "class GATv2(torch.nn.Module):\n",
+    "  def __init__(self,dim_h,heads=4):\n",
+    "    super().__init__()\n",
+    "    torch.manual_seed(55)\n",
+    "    edge_dim = dataset[0].edge_attr.shape[1]\n",
+    "    self.gat1=GATv2Conv(dataset.num_node_features,dim_h,heads=heads,concat=True,edge_dim=edge_dim)\n",
+    "    self.gat2=GATv2Conv(dim_h*heads,dim_h,heads=heads,concat=True,edge_dim=edge_dim)\n",
+    "    self.lin=torch.nn.Linear(dim_h*heads,dataset.num_classes)\n",
+    "  def forward(self,data):\n",
+    "    x,edge_index,edge_attr,batch=data.x,data.edge_index,data.edge_attr,data.batch\n",
+    "    x=self.gat1(x,edge_index,edge_attr)\n",
+    "    x=F.elu(x)\n",
+    "    x=self.gat2(x,edge_index,edge_attr)\n",
+    "    x=F.elu(x)\n",
+    "    x=global_mean_pool(x,batch)\n",
+    "    x=F.dropout(x,0.3,training=self.training)\n",
+    "    x=self.lin(x)\n",
+    "    return x\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 16,
+   "id": "58bb1ac9",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "dataset = dataset.shuffle()\n",
+    "train_size = int(0.8 * len(dataset))\n",
+    "test_size = len(dataset) - train_size\n",
+    "train_data,test_data=dataset[:train_size],dataset[train_size:]\n",
+    "train_loader = DataLoader(train_data, batch_size=32, shuffle=True)\n",
+    "test_loader = DataLoader(test_data, batch_size=32, shuffle=False)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 17,
+   "id": "2d6479c6",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch: 001, Train Acc: 0.4957, Test Acc: 0.4900, Time: 25.8774 seconds\n",
+      "Epoch: 002, Train Acc: 0.6084, Test Acc: 0.6093, Time: 24.4807 seconds\n",
+      "Epoch: 003, Train Acc: 0.6459, Test Acc: 0.6454, Time: 24.6017 seconds\n",
+      "Epoch: 004, Train Acc: 0.6813, Test Acc: 0.6749, Time: 24.6508 seconds\n",
+      "Epoch: 005, Train Acc: 0.6887, Test Acc: 0.6850, Time: 24.8438 seconds\n",
+      "Epoch: 006, Train Acc: 0.6733, Test Acc: 0.6679, Time: 24.9810 seconds\n",
+      "Epoch: 007, Train Acc: 0.7105, Test Acc: 0.7028, Time: 25.4978 seconds\n",
+      "Epoch: 008, Train Acc: 0.7156, Test Acc: 0.7070, Time: 26.5686 seconds\n",
+      "Epoch: 009, Train Acc: 0.7192, Test Acc: 0.7057, Time: 25.5068 seconds\n",
+      "Epoch: 010, Train Acc: 0.7452, Test Acc: 0.7364, Time: 24.8775 seconds\n",
+      "Epoch: 011, Train Acc: 0.7473, Test Acc: 0.7358, Time: 24.8751 seconds\n",
+      "Epoch: 012, Train Acc: 0.7590, Test Acc: 0.7523, Time: 24.8378 seconds\n",
+      "Epoch: 013, Train Acc: 0.7534, Test Acc: 0.7441, Time: 24.8861 seconds\n",
+      "Epoch: 014, Train Acc: 0.7390, Test Acc: 0.7297, Time: 24.8664 seconds\n",
+      "Epoch: 015, Train Acc: 0.7719, Test Acc: 0.7642, Time: 24.7913 seconds\n",
+      "Epoch: 016, Train Acc: 0.7905, Test Acc: 0.7778, Time: 24.7148 seconds\n",
+      "Epoch: 017, Train Acc: 0.7631, Test Acc: 0.7585, Time: 24.8516 seconds\n",
+      "Epoch: 018, Train Acc: 0.7694, Test Acc: 0.7562, Time: 24.8459 seconds\n",
+      "Epoch: 019, Train Acc: 0.7915, Test Acc: 0.7809, Time: 25.7345 seconds\n",
+      "Epoch: 020, Train Acc: 0.7866, Test Acc: 0.7721, Time: 25.8023 seconds\n",
+      "Epoch: 021, Train Acc: 0.7834, Test Acc: 0.7715, Time: 26.0532 seconds\n",
+      "Epoch: 022, Train Acc: 0.7917, Test Acc: 0.7799, Time: 25.6241 seconds\n",
+      "Epoch: 023, Train Acc: 0.7874, Test Acc: 0.7795, Time: 25.5449 seconds\n",
+      "Epoch: 024, Train Acc: 0.8107, Test Acc: 0.7905, Time: 25.6376 seconds\n",
+      "Epoch: 025, Train Acc: 0.8006, Test Acc: 0.7877, Time: 25.4805 seconds\n",
+      "Epoch: 026, Train Acc: 0.8106, Test Acc: 0.7940, Time: 25.5074 seconds\n",
+      "Epoch: 027, Train Acc: 0.8090, Test Acc: 0.7974, Time: 25.8159 seconds\n",
+      "Epoch: 028, Train Acc: 0.7693, Test Acc: 0.7540, Time: 25.4768 seconds\n",
+      "Epoch: 029, Train Acc: 0.8152, Test Acc: 0.7982, Time: 25.3734 seconds\n",
+      "Epoch: 030, Train Acc: 0.8120, Test Acc: 0.7951, Time: 25.2074 seconds\n",
+      "Epoch: 031, Train Acc: 0.7903, Test Acc: 0.7717, Time: 25.2981 seconds\n",
+      "Epoch: 032, Train Acc: 0.7993, Test Acc: 0.7815, Time: 25.2922 seconds\n",
+      "Epoch: 033, Train Acc: 0.8092, Test Acc: 0.7857, Time: 25.2716 seconds\n",
+      "Epoch: 034, Train Acc: 0.8086, Test Acc: 0.7821, Time: 25.1796 seconds\n",
+      "Epoch: 035, Train Acc: 0.8158, Test Acc: 0.7929, Time: 25.2028 seconds\n",
+      "Epoch: 036, Train Acc: 0.8133, Test Acc: 0.7926, Time: 25.1788 seconds\n",
+      "Epoch: 037, Train Acc: 0.8030, Test Acc: 0.7852, Time: 25.1870 seconds\n",
+      "Epoch: 038, Train Acc: 0.8300, Test Acc: 0.8062, Time: 25.2557 seconds\n",
+      "Epoch: 039, Train Acc: 0.8002, Test Acc: 0.7762, Time: 25.2280 seconds\n",
+      "Epoch: 040, Train Acc: 0.8199, Test Acc: 0.7973, Time: 25.2002 seconds\n",
+      "Epoch: 041, Train Acc: 0.8112, Test Acc: 0.7879, Time: 25.1714 seconds\n",
+      "Epoch: 042, Train Acc: 0.8256, Test Acc: 0.8116, Time: 25.2478 seconds\n",
+      "Epoch: 043, Train Acc: 0.8447, Test Acc: 0.8215, Time: 25.3352 seconds\n",
+      "Epoch: 044, Train Acc: 0.8315, Test Acc: 0.8033, Time: 26.1272 seconds\n",
+      "Epoch: 045, Train Acc: 0.8426, Test Acc: 0.8194, Time: 25.4368 seconds\n",
+      "Epoch: 046, Train Acc: 0.8236, Test Acc: 0.7975, Time: 25.2172 seconds\n",
+      "Epoch: 047, Train Acc: 0.8340, Test Acc: 0.8093, Time: 25.4511 seconds\n",
+      "Epoch: 048, Train Acc: 0.8483, Test Acc: 0.8281, Time: 25.2266 seconds\n",
+      "Epoch: 049, Train Acc: 0.8475, Test Acc: 0.8275, Time: 25.1697 seconds\n",
+      "TOtal time taken:1238.5092 seconds\n"
+     ]
+    }
+   ],
+   "source": [
+    "\n",
+    "model = GATv2(128).to(device)\n",
+    "optimizer = torch.optim.Adam(model.parameters(), lr=0.001)\n",
+    "criterion = torch.nn.CrossEntropyLoss()\n",
+    "\n",
+    "def train():\n",
+    "    model.train()\n",
+    "\n",
+    "    for data in train_loader:  # Iterate in batches over the training dataset.\n",
+    "         out = model(data.to(device))  # Perform a single forward pass.\n",
+    "         loss = criterion(out, data.y)  # Compute the loss.\n",
+    "         loss.backward()  # Derive gradients.\n",
+    "         optimizer.step()  # Update parameters based on gradients.\n",
+    "         optimizer.zero_grad()  # Clear gradients.\n",
+    "\n",
+    "def test(loader):\n",
+    "     model.eval()\n",
+    "\n",
+    "     correct = 0\n",
+    "     for data in loader:  # Iterate in batches over the training/test dataset.\n",
+    "         out = model(data.to(device))\n",
+    "         pred = out.argmax(dim=1)  # Use the class with highest probability.\n",
+    "         correct += int((pred == data.y).sum())  # Check against ground-truth labels.\n",
+    "     return correct / len(loader.dataset)  # Derive ratio of correct predictions.\n",
+    "\n",
+    "\n",
+    "acc,t=[],[]\n",
+    "a=time.time()\n",
+    "for epoch in range(1, 50):\n",
+    "    x=time.time()\n",
+    "    train()\n",
+    "    train_acc = test(train_loader)\n",
+    "    test_acc = test(test_loader)\n",
+    "    y=time.time()\n",
+    "    acc.append(test_acc)\n",
+    "    t.append(y-x)\n",
+    "    print(f'Epoch: {epoch:03d}, Train Acc: {train_acc:.4f}, Test Acc: {test_acc:.4f}, Time: {(y-x):.4f} seconds')\n",
+    "b=time.time()\n",
+    "print(f'TOtal time taken:{(b-a):.4f} seconds')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 18,
+   "id": "2d9dbd7c",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch: 001, Train Acc: 0.8265, Test Acc: 0.7945, Time: 24.9625 seconds\n",
+      "Epoch: 002, Train Acc: 0.8503, Test Acc: 0.8258, Time: 24.9976 seconds\n",
+      "Epoch: 003, Train Acc: 0.8506, Test Acc: 0.8281, Time: 25.1676 seconds\n",
+      "Epoch: 004, Train Acc: 0.8372, Test Acc: 0.8184, Time: 25.4236 seconds\n",
+      "Epoch: 005, Train Acc: 0.8557, Test Acc: 0.8248, Time: 27.4433 seconds\n",
+      "Epoch: 006, Train Acc: 0.8488, Test Acc: 0.8201, Time: 26.2164 seconds\n",
+      "Epoch: 007, Train Acc: 0.8493, Test Acc: 0.8219, Time: 28.7789 seconds\n",
+      "Epoch: 008, Train Acc: 0.8429, Test Acc: 0.8165, Time: 25.7886 seconds\n",
+      "Epoch: 009, Train Acc: 0.8613, Test Acc: 0.8276, Time: 26.8074 seconds\n",
+      "Epoch: 010, Train Acc: 0.8528, Test Acc: 0.8196, Time: 26.9662 seconds\n",
+      "Epoch: 011, Train Acc: 0.8372, Test Acc: 0.8053, Time: 25.6793 seconds\n",
+      "Epoch: 012, Train Acc: 0.8352, Test Acc: 0.8151, Time: 25.4075 seconds\n",
+      "Epoch: 013, Train Acc: 0.8603, Test Acc: 0.8337, Time: 25.2606 seconds\n",
+      "Epoch: 014, Train Acc: 0.8521, Test Acc: 0.8254, Time: 25.2944 seconds\n",
+      "Epoch: 015, Train Acc: 0.8498, Test Acc: 0.8143, Time: 25.2541 seconds\n",
+      "Epoch: 016, Train Acc: 0.8547, Test Acc: 0.8205, Time: 25.2576 seconds\n",
+      "Epoch: 017, Train Acc: 0.8482, Test Acc: 0.8220, Time: 25.1988 seconds\n",
+      "Epoch: 018, Train Acc: 0.8525, Test Acc: 0.8296, Time: 25.2256 seconds\n",
+      "Epoch: 019, Train Acc: 0.8595, Test Acc: 0.8259, Time: 25.4156 seconds\n",
+      "Epoch: 020, Train Acc: 0.8589, Test Acc: 0.8261, Time: 25.2268 seconds\n",
+      "Epoch: 021, Train Acc: 0.8658, Test Acc: 0.8304, Time: 25.4441 seconds\n",
+      "Epoch: 022, Train Acc: 0.8352, Test Acc: 0.8070, Time: 25.5956 seconds\n",
+      "Epoch: 023, Train Acc: 0.8724, Test Acc: 0.8299, Time: 25.4439 seconds\n",
+      "Epoch: 024, Train Acc: 0.8242, Test Acc: 0.7918, Time: 25.2510 seconds\n",
+      "Epoch: 025, Train Acc: 0.8627, Test Acc: 0.8307, Time: 26.1188 seconds\n",
+      "Epoch: 026, Train Acc: 0.8717, Test Acc: 0.8348, Time: 26.0500 seconds\n",
+      "Epoch: 027, Train Acc: 0.8544, Test Acc: 0.8184, Time: 25.5288 seconds\n",
+      "Epoch: 028, Train Acc: 0.7814, Test Acc: 0.7487, Time: 25.7745 seconds\n",
+      "Epoch: 029, Train Acc: 0.8569, Test Acc: 0.8292, Time: 25.3209 seconds\n",
+      "Epoch: 030, Train Acc: 0.8623, Test Acc: 0.8292, Time: 25.2030 seconds\n",
+      "Epoch: 031, Train Acc: 0.8640, Test Acc: 0.8287, Time: 25.9134 seconds\n",
+      "Epoch: 032, Train Acc: 0.8660, Test Acc: 0.8343, Time: 25.8696 seconds\n",
+      "Epoch: 033, Train Acc: 0.8541, Test Acc: 0.8167, Time: 25.1739 seconds\n",
+      "Epoch: 034, Train Acc: 0.8516, Test Acc: 0.8204, Time: 25.2224 seconds\n",
+      "Epoch: 035, Train Acc: 0.8651, Test Acc: 0.8274, Time: 25.4136 seconds\n",
+      "Epoch: 036, Train Acc: 0.8691, Test Acc: 0.8329, Time: 25.2840 seconds\n",
+      "Epoch: 037, Train Acc: 0.8794, Test Acc: 0.8351, Time: 25.4689 seconds\n",
+      "Epoch: 038, Train Acc: 0.8390, Test Acc: 0.8005, Time: 25.4441 seconds\n",
+      "Epoch: 039, Train Acc: 0.8515, Test Acc: 0.8180, Time: 25.4249 seconds\n",
+      "Epoch: 040, Train Acc: 0.8535, Test Acc: 0.8221, Time: 25.4681 seconds\n",
+      "Epoch: 041, Train Acc: 0.8727, Test Acc: 0.8391, Time: 25.1276 seconds\n",
+      "Epoch: 042, Train Acc: 0.8796, Test Acc: 0.8400, Time: 24.9614 seconds\n",
+      "Epoch: 043, Train Acc: 0.8708, Test Acc: 0.8330, Time: 25.3893 seconds\n",
+      "Epoch: 044, Train Acc: 0.8717, Test Acc: 0.8348, Time: 25.6369 seconds\n",
+      "Epoch: 045, Train Acc: 0.8654, Test Acc: 0.8273, Time: 25.1120 seconds\n",
+      "Epoch: 046, Train Acc: 0.8614, Test Acc: 0.8251, Time: 25.0642 seconds\n",
+      "Epoch: 047, Train Acc: 0.8609, Test Acc: 0.8207, Time: 25.0014 seconds\n",
+      "Epoch: 048, Train Acc: 0.8761, Test Acc: 0.8411, Time: 25.0245 seconds\n",
+      "Epoch: 049, Train Acc: 0.8633, Test Acc: 0.8239, Time: 25.0614 seconds\n",
+      "TOtal time taken:2581.8546 seconds\n"
+     ]
+    }
+   ],
+   "source": [
+    "for epoch in range(1, 50):\n",
+    "    x=time.time()\n",
+    "    train()\n",
+    "    train_acc = test(train_loader)\n",
+    "    test_acc = test(test_loader)\n",
+    "    y=time.time()\n",
+    "    acc.append(test_acc)\n",
+    "    t.append(y-x)\n",
+    "    print(f'Epoch: {epoch:03d}, Train Acc: {train_acc:.4f}, Test Acc: {test_acc:.4f}, Time: {(y-x):.4f} seconds')\n",
+    "b=time.time()\n",
+    "print(f'TOtal time taken:{(b-a):.4f} seconds')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 19,
+   "id": "d7c43502",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch: 001, Train Acc: 0.8673, Test Acc: 0.8311, Time: 25.3796 seconds\n",
+      "Epoch: 002, Train Acc: 0.8462, Test Acc: 0.8125, Time: 25.1563 seconds\n",
+      "Epoch: 003, Train Acc: 0.8498, Test Acc: 0.8034, Time: 25.2885 seconds\n",
+      "Epoch: 004, Train Acc: 0.8368, Test Acc: 0.8027, Time: 25.2527 seconds\n",
+      "Epoch: 005, Train Acc: 0.8805, Test Acc: 0.8356, Time: 26.5682 seconds\n",
+      "Epoch: 006, Train Acc: 0.8526, Test Acc: 0.8156, Time: 25.8687 seconds\n",
+      "Epoch: 007, Train Acc: 0.8837, Test Acc: 0.8427, Time: 25.6244 seconds\n",
+      "Epoch: 008, Train Acc: 0.8823, Test Acc: 0.8358, Time: 25.2722 seconds\n",
+      "Epoch: 009, Train Acc: 0.8740, Test Acc: 0.8337, Time: 25.2649 seconds\n",
+      "Epoch: 010, Train Acc: 0.8775, Test Acc: 0.8349, Time: 25.5523 seconds\n",
+      "Epoch: 011, Train Acc: 0.8671, Test Acc: 0.8299, Time: 25.1311 seconds\n",
+      "Epoch: 012, Train Acc: 0.8681, Test Acc: 0.8315, Time: 25.1274 seconds\n",
+      "Epoch: 013, Train Acc: 0.8867, Test Acc: 0.8430, Time: 25.1428 seconds\n",
+      "Epoch: 014, Train Acc: 0.8808, Test Acc: 0.8370, Time: 25.3863 seconds\n",
+      "Epoch: 015, Train Acc: 0.8705, Test Acc: 0.8361, Time: 25.3820 seconds\n",
+      "Epoch: 016, Train Acc: 0.8663, Test Acc: 0.8262, Time: 25.3648 seconds\n",
+      "Epoch: 017, Train Acc: 0.8764, Test Acc: 0.8315, Time: 25.1957 seconds\n",
+      "Epoch: 018, Train Acc: 0.8839, Test Acc: 0.8341, Time: 25.1296 seconds\n",
+      "Epoch: 019, Train Acc: 0.8744, Test Acc: 0.8259, Time: 25.1842 seconds\n",
+      "Epoch: 020, Train Acc: 0.8728, Test Acc: 0.8272, Time: 25.1960 seconds\n",
+      "Epoch: 021, Train Acc: 0.8814, Test Acc: 0.8411, Time: 25.2335 seconds\n",
+      "Epoch: 022, Train Acc: 0.8612, Test Acc: 0.8265, Time: 25.2171 seconds\n",
+      "Epoch: 023, Train Acc: 0.8821, Test Acc: 0.8402, Time: 25.2572 seconds\n",
+      "Epoch: 024, Train Acc: 0.8830, Test Acc: 0.8370, Time: 26.0306 seconds\n",
+      "Epoch: 025, Train Acc: 0.8693, Test Acc: 0.8261, Time: 25.1953 seconds\n",
+      "Epoch: 026, Train Acc: 0.8698, Test Acc: 0.8288, Time: 25.1328 seconds\n",
+      "Epoch: 027, Train Acc: 0.8853, Test Acc: 0.8447, Time: 25.1489 seconds\n",
+      "Epoch: 028, Train Acc: 0.8785, Test Acc: 0.8285, Time: 25.2630 seconds\n",
+      "Epoch: 029, Train Acc: 0.8882, Test Acc: 0.8385, Time: 25.1945 seconds\n",
+      "Epoch: 030, Train Acc: 0.8726, Test Acc: 0.8288, Time: 25.1916 seconds\n",
+      "Epoch: 031, Train Acc: 0.8895, Test Acc: 0.8386, Time: 25.2485 seconds\n",
+      "Epoch: 032, Train Acc: 0.8636, Test Acc: 0.8295, Time: 25.1767 seconds\n",
+      "Epoch: 033, Train Acc: 0.8746, Test Acc: 0.8299, Time: 25.2032 seconds\n",
+      "Epoch: 034, Train Acc: 0.8880, Test Acc: 0.8383, Time: 25.5541 seconds\n",
+      "Epoch: 035, Train Acc: 0.8822, Test Acc: 0.8385, Time: 25.2495 seconds\n",
+      "Epoch: 036, Train Acc: 0.8909, Test Acc: 0.8467, Time: 25.4243 seconds\n",
+      "Epoch: 037, Train Acc: 0.8839, Test Acc: 0.8351, Time: 25.5948 seconds\n",
+      "Epoch: 038, Train Acc: 0.8821, Test Acc: 0.8341, Time: 25.2591 seconds\n",
+      "Epoch: 039, Train Acc: 0.8889, Test Acc: 0.8354, Time: 25.2509 seconds\n",
+      "Epoch: 040, Train Acc: 0.8842, Test Acc: 0.8446, Time: 25.2162 seconds\n",
+      "Epoch: 041, Train Acc: 0.8636, Test Acc: 0.8126, Time: 25.2191 seconds\n",
+      "Epoch: 042, Train Acc: 0.8817, Test Acc: 0.8363, Time: 25.2106 seconds\n",
+      "Epoch: 043, Train Acc: 0.8897, Test Acc: 0.8368, Time: 25.2143 seconds\n",
+      "Epoch: 044, Train Acc: 0.8777, Test Acc: 0.8316, Time: 25.1635 seconds\n",
+      "Epoch: 045, Train Acc: 0.8879, Test Acc: 0.8396, Time: 25.2399 seconds\n",
+      "Epoch: 046, Train Acc: 0.8850, Test Acc: 0.8309, Time: 25.1394 seconds\n",
+      "Epoch: 047, Train Acc: 0.8801, Test Acc: 0.8281, Time: 25.3154 seconds\n",
+      "Epoch: 048, Train Acc: 0.8905, Test Acc: 0.8363, Time: 25.1571 seconds\n",
+      "Epoch: 049, Train Acc: 0.8916, Test Acc: 0.8450, Time: 25.1891 seconds\n",
+      "TOtal time taken:4324.3672 seconds\n"
+     ]
+    }
+   ],
+   "source": [
+    "for epoch in range(1, 50):\n",
+    "    x=time.time()\n",
+    "    train()\n",
+    "    train_acc = test(train_loader)\n",
+    "    test_acc = test(test_loader)\n",
+    "    y=time.time()\n",
+    "    acc.append(test_acc)\n",
+    "    t.append(y-x)\n",
+    "    print(f'Epoch: {epoch:03d}, Train Acc: {train_acc:.4f}, Test Acc: {test_acc:.4f}, Time: {(y-x):.4f} seconds')\n",
+    "b=time.time()\n",
+    "print(f'TOtal time taken:{(b-a):.4f} seconds')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 20,
+   "id": "ccdc2139",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import torch\n",
+    "import torch.nn as nn\n",
+    "import torch.nn.functional as F\n",
+    "from torch_geometric.nn import TransformerConv, global_mean_pool\n",
+    "\n",
+    "class GraphTransformerNet(nn.Module):\n",
+    "    def __init__(self, in_channels, hidden_channels, out_channels, heads=4, num_layers=3, dropout=0.2):\n",
+    "        super().__init__()\n",
+    "        self.convs = nn.ModuleList()\n",
+    "        self.convs.append(TransformerConv(in_channels, hidden_channels, heads=heads, dropout=dropout))\n",
+    "        for _ in range(num_layers - 2):\n",
+    "            self.convs.append(TransformerConv(hidden_channels * heads, hidden_channels, heads=heads, dropout=dropout))\n",
+    "        self.convs.append(TransformerConv(hidden_channels * heads, hidden_channels, heads=1, dropout=dropout))\n",
+    "        self.lin = nn.Linear(hidden_channels, out_channels)\n",
+    "        self.dropout = dropout\n",
+    "\n",
+    "    def forward(self, data):\n",
+    "        x, edge_index, batch = data.x, data.edge_index, data.batch\n",
+    "        for conv in self.convs:\n",
+    "            x = conv(x, edge_index)\n",
+    "            x = F.relu(x)\n",
+    "            x = F.dropout(x, p=self.dropout, training=self.training)\n",
+    "        x = global_mean_pool(x, batch)\n",
+    "        x = self.lin(x)\n",
+    "        return x"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 26,
+   "id": "66fc81fa",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch: 001, Train Acc: 0.4981, Test Acc: 0.5008, Time: 21.0400 seconds\n",
+      "Epoch: 002, Train Acc: 0.6214, Test Acc: 0.6283, Time: 21.0401 seconds\n",
+      "Epoch: 003, Train Acc: 0.6922, Test Acc: 0.6972, Time: 21.3110 seconds\n",
+      "Epoch: 004, Train Acc: 0.6340, Test Acc: 0.6347, Time: 21.4084 seconds\n",
+      "Epoch: 005, Train Acc: 0.7140, Test Acc: 0.7116, Time: 21.5427 seconds\n",
+      "Epoch: 006, Train Acc: 0.7017, Test Acc: 0.6966, Time: 21.5673 seconds\n",
+      "Epoch: 007, Train Acc: 0.7422, Test Acc: 0.7420, Time: 21.4987 seconds\n",
+      "Epoch: 008, Train Acc: 0.7296, Test Acc: 0.7319, Time: 21.3965 seconds\n",
+      "Epoch: 009, Train Acc: 0.7007, Test Acc: 0.7085, Time: 21.3689 seconds\n",
+      "Epoch: 010, Train Acc: 0.7653, Test Acc: 0.7651, Time: 21.3604 seconds\n",
+      "Epoch: 011, Train Acc: 0.7580, Test Acc: 0.7622, Time: 21.3570 seconds\n",
+      "Epoch: 012, Train Acc: 0.7195, Test Acc: 0.7173, Time: 21.3565 seconds\n",
+      "Epoch: 013, Train Acc: 0.7629, Test Acc: 0.7655, Time: 21.4036 seconds\n",
+      "Epoch: 014, Train Acc: 0.7690, Test Acc: 0.7669, Time: 21.3677 seconds\n",
+      "Epoch: 015, Train Acc: 0.7581, Test Acc: 0.7540, Time: 21.4620 seconds\n",
+      "Epoch: 016, Train Acc: 0.7948, Test Acc: 0.7952, Time: 21.4100 seconds\n",
+      "Epoch: 017, Train Acc: 0.7715, Test Acc: 0.7711, Time: 21.5848 seconds\n",
+      "Epoch: 018, Train Acc: 0.7848, Test Acc: 0.7813, Time: 21.4893 seconds\n",
+      "Epoch: 019, Train Acc: 0.8042, Test Acc: 0.7989, Time: 21.5196 seconds\n",
+      "Epoch: 020, Train Acc: 0.7848, Test Acc: 0.7852, Time: 21.4080 seconds\n",
+      "Epoch: 021, Train Acc: 0.7854, Test Acc: 0.7845, Time: 21.3282 seconds\n",
+      "Epoch: 022, Train Acc: 0.7995, Test Acc: 0.7979, Time: 21.3296 seconds\n",
+      "Epoch: 023, Train Acc: 0.8046, Test Acc: 0.8013, Time: 21.2942 seconds\n",
+      "Epoch: 024, Train Acc: 0.7934, Test Acc: 0.7945, Time: 21.2278 seconds\n",
+      "Epoch: 025, Train Acc: 0.7872, Test Acc: 0.7893, Time: 21.3140 seconds\n",
+      "Epoch: 026, Train Acc: 0.8161, Test Acc: 0.8152, Time: 21.2853 seconds\n",
+      "Epoch: 027, Train Acc: 0.8057, Test Acc: 0.8043, Time: 21.2965 seconds\n",
+      "Epoch: 028, Train Acc: 0.8098, Test Acc: 0.8067, Time: 21.2929 seconds\n",
+      "Epoch: 029, Train Acc: 0.8096, Test Acc: 0.8034, Time: 21.3359 seconds\n",
+      "TOtal time taken:619.6132 seconds\n"
+     ]
+    }
+   ],
+   "source": [
+    "in_channels = dataset.num_node_features\n",
+    "hidden_channels = 128\n",
+    "out_channels = dataset.num_classes\n",
+    "model = GraphTransformerNet(in_channels, hidden_channels, out_channels).to(device)\n",
+    "optimizer = torch.optim.Adam(model.parameters(), lr=0.001)\n",
+    "criterion = torch.nn.CrossEntropyLoss()\n",
+    "\n",
+    "def train():\n",
+    "    model.train()\n",
+    "\n",
+    "    for data in train_loader:  # Iterate in batches over the training dataset.\n",
+    "         out = model(data.to(device))  # Perform a single forward pass.\n",
+    "         loss = criterion(out, data.y)  # Compute the loss.\n",
+    "         loss.backward()  # Derive gradients.\n",
+    "         optimizer.step()  # Update parameters based on gradients.\n",
+    "         optimizer.zero_grad()  # Clear gradients.\n",
+    "\n",
+    "def test(loader):\n",
+    "     model.eval()\n",
+    "\n",
+    "     correct = 0\n",
+    "     for data in loader:  # Iterate in batches over the training/test dataset.\n",
+    "         out = model(data.to(device))\n",
+    "         pred = out.argmax(dim=1)  # Use the class with highest probability.\n",
+    "         correct += int((pred == data.y).sum())  # Check against ground-truth labels.\n",
+    "     return correct / len(loader.dataset)  # Derive ratio of correct predictions.\n",
+    "\n",
+    "\n",
+    "acc,t=[],[]\n",
+    "a=time.time()\n",
+    "for epoch in range(1, 30):\n",
+    "    x=time.time()\n",
+    "    train()\n",
+    "    train_acc = test(train_loader)\n",
+    "    test_acc = test(test_loader)\n",
+    "    y=time.time()\n",
+    "    acc.append(test_acc)\n",
+    "    t.append(y-x)\n",
+    "    print(f'Epoch: {epoch:03d}, Train Acc: {train_acc:.4f}, Test Acc: {test_acc:.4f}, Time: {(y-x):.4f} seconds')\n",
+    "b=time.time()\n",
+    "print(f'TOtal time taken:{(b-a):.4f} seconds')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 27,
+   "id": "e160cdc1",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch: 001, Train Acc: 0.7834, Test Acc: 0.7799, Time: 20.5345 seconds\n",
+      "Epoch: 002, Train Acc: 0.8031, Test Acc: 0.7999, Time: 20.3483 seconds\n",
+      "Epoch: 003, Train Acc: 0.8063, Test Acc: 0.8047, Time: 20.4774 seconds\n",
+      "Epoch: 004, Train Acc: 0.7843, Test Acc: 0.7820, Time: 20.6039 seconds\n",
+      "Epoch: 005, Train Acc: 0.8096, Test Acc: 0.8084, Time: 20.7035 seconds\n",
+      "Epoch: 006, Train Acc: 0.8012, Test Acc: 0.7979, Time: 20.8724 seconds\n",
+      "Epoch: 007, Train Acc: 0.8301, Test Acc: 0.8248, Time: 20.9738 seconds\n",
+      "Epoch: 008, Train Acc: 0.7965, Test Acc: 0.7953, Time: 21.0458 seconds\n",
+      "Epoch: 009, Train Acc: 0.8086, Test Acc: 0.8059, Time: 21.1957 seconds\n",
+      "Epoch: 010, Train Acc: 0.8270, Test Acc: 0.8240, Time: 21.0377 seconds\n",
+      "Epoch: 011, Train Acc: 0.8156, Test Acc: 0.8117, Time: 21.0017 seconds\n",
+      "Epoch: 012, Train Acc: 0.8316, Test Acc: 0.8289, Time: 21.0952 seconds\n",
+      "Epoch: 013, Train Acc: 0.7880, Test Acc: 0.7836, Time: 21.0399 seconds\n",
+      "Epoch: 014, Train Acc: 0.8383, Test Acc: 0.8301, Time: 21.0658 seconds\n",
+      "Epoch: 015, Train Acc: 0.8307, Test Acc: 0.8258, Time: 21.0292 seconds\n",
+      "Epoch: 016, Train Acc: 0.8166, Test Acc: 0.8103, Time: 21.0862 seconds\n",
+      "Epoch: 017, Train Acc: 0.8419, Test Acc: 0.8398, Time: 21.0714 seconds\n",
+      "Epoch: 018, Train Acc: 0.8011, Test Acc: 0.7987, Time: 21.1226 seconds\n",
+      "Epoch: 019, Train Acc: 0.8289, Test Acc: 0.8227, Time: 21.1413 seconds\n",
+      "Epoch: 020, Train Acc: 0.8518, Test Acc: 0.8485, Time: 21.1090 seconds\n",
+      "Epoch: 021, Train Acc: 0.8334, Test Acc: 0.8294, Time: 21.0528 seconds\n",
+      "Epoch: 022, Train Acc: 0.8120, Test Acc: 0.8129, Time: 21.0969 seconds\n",
+      "Epoch: 023, Train Acc: 0.7796, Test Acc: 0.7753, Time: 21.0279 seconds\n",
+      "Epoch: 024, Train Acc: 0.8334, Test Acc: 0.8309, Time: 21.0795 seconds\n",
+      "Epoch: 025, Train Acc: 0.8322, Test Acc: 0.8260, Time: 21.1273 seconds\n",
+      "Epoch: 026, Train Acc: 0.8368, Test Acc: 0.8294, Time: 21.1004 seconds\n",
+      "Epoch: 027, Train Acc: 0.8486, Test Acc: 0.8403, Time: 21.0755 seconds\n",
+      "Epoch: 028, Train Acc: 0.8426, Test Acc: 0.8352, Time: 21.1000 seconds\n",
+      "Epoch: 029, Train Acc: 0.8409, Test Acc: 0.8317, Time: 21.0839 seconds\n",
+      "Epoch: 030, Train Acc: 0.8315, Test Acc: 0.8276, Time: 21.0667 seconds\n",
+      "Epoch: 031, Train Acc: 0.8310, Test Acc: 0.8259, Time: 21.0606 seconds\n",
+      "Epoch: 032, Train Acc: 0.8423, Test Acc: 0.8378, Time: 21.0881 seconds\n",
+      "Epoch: 033, Train Acc: 0.8564, Test Acc: 0.8525, Time: 21.0627 seconds\n",
+      "Epoch: 034, Train Acc: 0.8418, Test Acc: 0.8372, Time: 21.2976 seconds\n",
+      "Epoch: 035, Train Acc: 0.8245, Test Acc: 0.8176, Time: 21.2908 seconds\n",
+      "Epoch: 036, Train Acc: 0.8377, Test Acc: 0.8315, Time: 21.3887 seconds\n",
+      "Epoch: 037, Train Acc: 0.8294, Test Acc: 0.8248, Time: 21.3492 seconds\n",
+      "Epoch: 038, Train Acc: 0.8311, Test Acc: 0.8249, Time: 21.3444 seconds\n",
+      "Epoch: 039, Train Acc: 0.8646, Test Acc: 0.8570, Time: 21.3087 seconds\n",
+      "Epoch: 040, Train Acc: 0.8524, Test Acc: 0.8457, Time: 21.3876 seconds\n",
+      "Epoch: 041, Train Acc: 0.8398, Test Acc: 0.8322, Time: 21.3600 seconds\n",
+      "Epoch: 042, Train Acc: 0.8227, Test Acc: 0.8204, Time: 21.3573 seconds\n",
+      "Epoch: 043, Train Acc: 0.8645, Test Acc: 0.8574, Time: 21.3399 seconds\n",
+      "Epoch: 044, Train Acc: 0.8458, Test Acc: 0.8381, Time: 21.3396 seconds\n",
+      "Epoch: 045, Train Acc: 0.8712, Test Acc: 0.8624, Time: 21.4001 seconds\n",
+      "Epoch: 046, Train Acc: 0.8596, Test Acc: 0.8474, Time: 21.3271 seconds\n",
+      "Epoch: 047, Train Acc: 0.8643, Test Acc: 0.8560, Time: 21.4970 seconds\n",
+      "Epoch: 048, Train Acc: 0.8583, Test Acc: 0.8511, Time: 21.3742 seconds\n",
+      "Epoch: 049, Train Acc: 0.8456, Test Acc: 0.8427, Time: 21.3583 seconds\n",
+      "TOtal time taken:3346.5106 seconds\n"
+     ]
+    }
+   ],
+   "source": [
+    "for epoch in range(1, 50):\n",
+    "    x=time.time()\n",
+    "    train()\n",
+    "    train_acc = test(train_loader)\n",
+    "    test_acc = test(test_loader)\n",
+    "    y=time.time()\n",
+    "    acc.append(test_acc)\n",
+    "    t.append(y-x)\n",
+    "    print(f'Epoch: {epoch:03d}, Train Acc: {train_acc:.4f}, Test Acc: {test_acc:.4f}, Time: {(y-x):.4f} seconds')\n",
+    "b=time.time()\n",
+    "print(f'TOtal time taken:{(b-a):.4f} seconds')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 28,
+   "id": "b5273211",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch: 001, Train Acc: 0.8294, Test Acc: 0.8244, Time: 21.6335 seconds\n",
+      "Epoch: 002, Train Acc: 0.8411, Test Acc: 0.8347, Time: 21.9815 seconds\n",
+      "Epoch: 003, Train Acc: 0.8580, Test Acc: 0.8525, Time: 22.3066 seconds\n",
+      "Epoch: 004, Train Acc: 0.8512, Test Acc: 0.8456, Time: 21.8491 seconds\n",
+      "Epoch: 005, Train Acc: 0.8688, Test Acc: 0.8576, Time: 21.3336 seconds\n",
+      "Epoch: 006, Train Acc: 0.8604, Test Acc: 0.8543, Time: 21.8262 seconds\n",
+      "Epoch: 007, Train Acc: 0.8311, Test Acc: 0.8282, Time: 21.4122 seconds\n",
+      "Epoch: 008, Train Acc: 0.8623, Test Acc: 0.8533, Time: 21.7136 seconds\n",
+      "Epoch: 009, Train Acc: 0.8545, Test Acc: 0.8516, Time: 21.7051 seconds\n",
+      "Epoch: 010, Train Acc: 0.8580, Test Acc: 0.8478, Time: 21.6664 seconds\n",
+      "Epoch: 011, Train Acc: 0.8400, Test Acc: 0.8350, Time: 22.2604 seconds\n",
+      "Epoch: 012, Train Acc: 0.8523, Test Acc: 0.8471, Time: 22.7480 seconds\n",
+      "Epoch: 013, Train Acc: 0.8662, Test Acc: 0.8597, Time: 21.5961 seconds\n",
+      "Epoch: 014, Train Acc: 0.8509, Test Acc: 0.8452, Time: 21.8072 seconds\n",
+      "Epoch: 015, Train Acc: 0.8710, Test Acc: 0.8648, Time: 22.6861 seconds\n",
+      "Epoch: 016, Train Acc: 0.8372, Test Acc: 0.8322, Time: 21.4229 seconds\n",
+      "Epoch: 017, Train Acc: 0.8435, Test Acc: 0.8354, Time: 21.4031 seconds\n",
+      "Epoch: 018, Train Acc: 0.8770, Test Acc: 0.8671, Time: 21.4051 seconds\n",
+      "Epoch: 019, Train Acc: 0.8349, Test Acc: 0.8281, Time: 21.3081 seconds\n",
+      "Epoch: 020, Train Acc: 0.8362, Test Acc: 0.8324, Time: 21.2875 seconds\n",
+      "Epoch: 021, Train Acc: 0.8589, Test Acc: 0.8511, Time: 21.6692 seconds\n",
+      "Epoch: 022, Train Acc: 0.8708, Test Acc: 0.8646, Time: 21.2015 seconds\n",
+      "Epoch: 023, Train Acc: 0.8649, Test Acc: 0.8575, Time: 21.3023 seconds\n",
+      "Epoch: 024, Train Acc: 0.8543, Test Acc: 0.8530, Time: 21.2712 seconds\n",
+      "Epoch: 025, Train Acc: 0.8542, Test Acc: 0.8491, Time: 21.0625 seconds\n",
+      "Epoch: 026, Train Acc: 0.8427, Test Acc: 0.8381, Time: 21.1436 seconds\n",
+      "Epoch: 027, Train Acc: 0.8594, Test Acc: 0.8523, Time: 21.2964 seconds\n",
+      "Epoch: 028, Train Acc: 0.8624, Test Acc: 0.8580, Time: 21.2751 seconds\n",
+      "Epoch: 029, Train Acc: 0.8614, Test Acc: 0.8568, Time: 21.3692 seconds\n",
+      "Epoch: 030, Train Acc: 0.8630, Test Acc: 0.8525, Time: 21.1976 seconds\n",
+      "Epoch: 031, Train Acc: 0.8545, Test Acc: 0.8501, Time: 21.0957 seconds\n",
+      "Epoch: 032, Train Acc: 0.8444, Test Acc: 0.8355, Time: 21.1121 seconds\n",
+      "Epoch: 033, Train Acc: 0.8571, Test Acc: 0.8521, Time: 21.9765 seconds\n",
+      "Epoch: 034, Train Acc: 0.8809, Test Acc: 0.8700, Time: 21.5728 seconds\n",
+      "Epoch: 035, Train Acc: 0.8607, Test Acc: 0.8502, Time: 21.7194 seconds\n",
+      "Epoch: 036, Train Acc: 0.8479, Test Acc: 0.8417, Time: 21.4821 seconds\n",
+      "Epoch: 037, Train Acc: 0.8483, Test Acc: 0.8434, Time: 21.6710 seconds\n",
+      "Epoch: 038, Train Acc: 0.8621, Test Acc: 0.8519, Time: 21.8791 seconds\n",
+      "Epoch: 039, Train Acc: 0.8502, Test Acc: 0.8463, Time: 21.6123 seconds\n",
+      "Epoch: 040, Train Acc: 0.8522, Test Acc: 0.8479, Time: 21.8715 seconds\n",
+      "Epoch: 041, Train Acc: 0.8525, Test Acc: 0.8450, Time: 21.4852 seconds\n",
+      "Epoch: 042, Train Acc: 0.8576, Test Acc: 0.8492, Time: 21.4420 seconds\n",
+      "Epoch: 043, Train Acc: 0.8381, Test Acc: 0.8347, Time: 21.7751 seconds\n",
+      "Epoch: 044, Train Acc: 0.8674, Test Acc: 0.8605, Time: 21.4741 seconds\n",
+      "Epoch: 045, Train Acc: 0.8588, Test Acc: 0.8540, Time: 21.4491 seconds\n",
+      "Epoch: 046, Train Acc: 0.8491, Test Acc: 0.8400, Time: 21.4769 seconds\n",
+      "Epoch: 047, Train Acc: 0.8439, Test Acc: 0.8376, Time: 21.4490 seconds\n",
+      "Epoch: 048, Train Acc: 0.8827, Test Acc: 0.8708, Time: 21.4509 seconds\n",
+      "Epoch: 049, Train Acc: 0.8777, Test Acc: 0.8652, Time: 21.4486 seconds\n",
+      "TOtal time taken:5775.3160 seconds\n"
+     ]
+    }
+   ],
+   "source": [
+    "for epoch in range(1, 50):\n",
+    "    x=time.time()\n",
+    "    train()\n",
+    "    train_acc = test(train_loader)\n",
+    "    test_acc = test(test_loader)\n",
+    "    y=time.time()\n",
+    "    acc.append(test_acc)\n",
+    "    t.append(y-x)\n",
+    "    print(f'Epoch: {epoch:03d}, Train Acc: {train_acc:.4f}, Test Acc: {test_acc:.4f}, Time: {(y-x):.4f} seconds')\n",
+    "b=time.time()\n",
+    "print(f'TOtal time taken:{(b-a):.4f} seconds')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "5052b422",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "for epoch in range(1, 50):\n",
+    "    x=time.time()\n",
+    "    train()\n",
+    "    train_acc = test(train_loader)\n",
+    "    test_acc = test(test_loader)\n",
+    "    y=time.time()\n",
+    "    acc.append(test_acc)\n",
+    "    t.append(y-x)\n",
+    "    print(f'Epoch: {epoch:03d}, Train Acc: {train_acc:.4f}, Test Acc: {test_acc:.4f}, Time: {(y-x):.4f} seconds')\n",
+    "b=time.time()\n",
+    "print(f'TOtal time taken:{(b-a):.4f} seconds')"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Project",
+   "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": "3.11.9"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

requirements.txt

@@ -0,0 +1,91 @@
+ase==3.25.0
+asttokens==3.0.0
+attrs==25.3.0
+catboost==1.2.8
+certifi==2025.1.31
+charset-normalizer==3.4.1
+colorama==0.4.6
+comm==0.2.2
+contourpy==1.3.2
+cycler==0.12.1
+debugpy==1.8.14
+decorator==5.2.1
+executing==2.2.0
+fastjsonschema==2.21.1
+filelock==3.13.1
+fonttools==4.57.0
+fsspec==2024.6.1
+-e git+https://github.com/rampasek/GraphGPS.git@28015707cbab7f8ad72bed0ee872d068ea59c94b#egg=graphgps
+graphviz==0.20.3
+idna==3.10
+ipykernel==6.29.5
+ipython==9.1.0
+ipython_pygments_lexers==1.1.1
+ipywidgets==8.1.6
+jedi==0.19.2
+Jinja2==3.1.4
+joblib==1.4.2
+jsonschema==4.23.0
+jsonschema-specifications==2025.4.1
+jupyter_client==8.6.3
+jupyter_core==5.7.2
+jupyterlab_widgets==3.0.14
+kiwisolver==1.4.8
+littleutils==0.2.4
+llvmlite==0.44.0
+MarkupSafe==2.1.5
+matplotlib==3.10.1
+matplotlib-inline==0.1.7
+mpmath==1.3.0
+narwhals==1.36.0
+nbformat==5.10.4
+nest-asyncio==1.6.0
+networkx==3.3
+numba==0.61.2
+numpy==2.2.5
+ogb==1.3.6
+outdated==0.2.2
+packaging==25.0
+pandas==2.2.3
+parso==0.8.4
+pillow==11.2.1
+platformdirs==4.3.7
+plotly==6.0.1
+prompt_toolkit==3.0.51
+psutil==7.0.0
+pure_eval==0.2.3
+Pygments==2.19.1
+pynndescent==0.5.13
+pynverse==0.1.4.6
+pyparsing==3.2.3
+python-dateutil==2.9.0.post0
+pytz==2025.2
+pywin32==310
+pyzmq==26.4.0
+rascal==0.3.10
+referencing==0.36.2
+requests==2.32.3
+rpds-py==0.24.0
+scikit-learn==1.6.1
+scipy==1.15.2
+seaborn==0.13.2
+six==1.17.0
+stack-data==0.6.3
+sympy==1.13.3
+threadpoolctl==3.6.0
+torch==2.4.1+cu121
+torch_cluster==1.6.3+pt24cu121
+torch_geometric==2.3.1
+torch_scatter==2.1.2+pt24cu121
+torch_sparse==0.6.18+pt24cu121
+torch_spline_conv==1.2.2+pt24cu121
+tornado==6.4.2
+tqdm==4.67.1
+traitlets==5.14.3
+typing_extensions==4.13.2
+tzdata==2025.2
+umap==0.1.1
+umap-learn==0.5.7
+urllib3==2.4.0
+wcwidth==0.2.13
+widgetsnbextension==4.0.14

san_best.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0dd7b736400736375f93e36fe2e5150e2c21258a709d998b028dd6900fd112fc
+size 757556

transformer_classifier_relu.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c76651ae52ee11de9167c281960117bd01c0d5b6a326dc5fef74caeff14ade83
+size 1540122