{
 "cells": [
  {
   "cell_type": "markdown",
   "source": [
    "# Train the Model"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "source": [
    "import torch\n",
    "from train_model import train_model\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "from torch.optim import lr_scheduler\n",
    "from model import build_hybrid_model\n",
    "import pennylane as qml\n",
    "\n",
    "STEP = 0.0004  # Learning rate\n",
    "BATCH_SIZE = 4  # Number of samples for each training step\n",
    "N_QUBITS = 4\n",
    "NUM_EPOCHS = 3  # Number of training epochs\n",
    "GAMMA_LR_SCHEDULER = 0.1  # Learning rate reduction applied every 10 epochs.\n",
    "PARAMETER_FILEPATH = \"state.pt\" # \n",
    "IF_INITIAL_TRAIN = False # Whether this is the first time training the model\n",
    "\n",
    "# Initialize Pennylane backend\n",
    "dev = qml.device(\"default.qubit\", wires=N_QUBITS)\n",
    "\n",
    "# Initialize PyTorch config\n",
    "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
    "\n",
    "# Load Model\n",
    "model = build_hybrid_model(\n",
    "    pennylane_dev=dev,\n",
    "    device=device,\n",
    "    n_qubits=N_QUBITS,\n",
    "    q_depth=6,\n",
    "    q_delta=0.01\n",
    ")\n",
    "\n",
    "if not IF_INITIAL_TRAIN:\n",
    "    # Load Model Weights from Disk\n",
    "    print(\"Loading Model Weights from Disk...\")\n",
    "    model.load_state_dict(torch.load(PARAMETER_FILEPATH))\n",
    "\n",
    "# Define Loss for Training\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "optimizer_hybrid = optim.Adam(model.fc.parameters(), lr=STEP)\n",
    "exp_lr_scheduler = lr_scheduler.StepLR(\n",
    "    optimizer_hybrid, step_size=10, gamma=GAMMA_LR_SCHEDULER\n",
    ")\n",
    "\n",
    "# Train the Model\n",
    "trained_model = train_model(\n",
    "    model=model, \n",
    "    criterion=criterion, \n",
    "    optimizer=optimizer_hybrid, \n",
    "    scheduler=exp_lr_scheduler, \n",
    "    num_epochs=NUM_EPOCHS,\n",
    "    batch_size=BATCH_SIZE,\n",
    "    device=device\n",
    ")\n",
    "\n",
    "# Save Model Parameters to Disk\n",
    "print(\"Saving Model Weights to Disk...\")\n",
    "torch.save(trained_model.state_dict(), PARAMETER_FILEPATH)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-09-11T22:38:45.705122Z",
     "start_time": "2024-09-11T22:16:02.896239Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\MaxHu\\anaconda3\\envs\\quantum-env\\lib\\site-packages\\torchvision\\models\\_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.\n",
      "  warnings.warn(\n",
      "C:\\Users\\MaxHu\\anaconda3\\envs\\quantum-env\\lib\\site-packages\\torchvision\\models\\_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=ResNet18_Weights.IMAGENET1K_V1`. You can also use `weights=ResNet18_Weights.DEFAULT` to get the most up-to-date weights.\n",
      "  warnings.warn(msg)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading Model Weights from Disk...\n",
      "Training started:\n",
      "Data Loading Completed in 0m 0s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\MaxHu\\AppData\\Local\\Temp\\ipykernel_12896\\449253314.py:35: 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",
      "  model.load_state_dict(torch.load(PARAMETER_FILEPATH))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Phase: train Epoch: 1/3 Loss: 0.8986 Acc: 0.5877        33\n",
      "Phase: validation   Epoch: 1/3 Loss: 1.9974 Acc: 0.3929        \n",
      "Phase: train Epoch: 2/3 Loss: 0.8994 Acc: 0.5849        76\n",
      "Phase: validation   Epoch: 2/3 Loss: 1.9718 Acc: 0.3929        \n",
      "Phase: train Epoch: 3/3 Loss: 0.8977 Acc: 0.5860        76\n",
      "Phase: validation   Epoch: 3/3 Loss: 2.0105 Acc: 0.3750        \n",
      "Training Completed in 22m 34s\n",
      "Best test loss: 1.9718 | Best test accuracy: 0.3929\n",
      "Saving Model Weights to Disk...\n"
     ]
    }
   ],
   "execution_count": 1
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Visualize a Batch of Predictions"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "source": [
    "from helpers import visualize_model\n",
    "import torch\n",
    "from train_model import train_model\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "from torch.optim import lr_scheduler\n",
    "from model import build_hybrid_model\n",
    "import pennylane as qml\n",
    "\n",
    "BATCH_SIZE = 4\n",
    "N_QUBITS = 4\n",
    "PARAMETER_FILEPATH = \"state.pt\"\n",
    "\n",
    "# Initialize Pennylane backend\n",
    "dev = qml.device(\"default.qubit\", wires=N_QUBITS)\n",
    "\n",
    "# Initialize PyTorch config\n",
    "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
    "\n",
    "# Load Model\n",
    "model = build_hybrid_model(\n",
    "    pennylane_dev=dev,\n",
    "    device=device,\n",
    "    n_qubits=N_QUBITS,\n",
    "    q_depth=6,\n",
    "    q_delta=0.01\n",
    ")\n",
    "\n",
    "# Load Model Weights from Disk\n",
    "print(\"Loading Model Weights from Disk...\")\n",
    "model.load_state_dict(torch.load(PARAMETER_FILEPATH))\n",
    "\n",
    "visualize_model(\n",
    "    model=model,\n",
    "    device=device,\n",
    "    batch_size=BATCH_SIZE\n",
    ")"
   ],
   "outputs": [],
   "execution_count": null
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Prediction Test"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "source": [
    "from generate_prediction import generate_prediction\n",
    "pred = generate_prediction(\"./image.jpeg\")\n",
    "print(\"PREDICTION:\",pred)"
   ],
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "execution_count": null
  },
  {
   "cell_type": "code",
   "source": "from app import *",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "execution_count": null
  },
  {
   "cell_type": "markdown",
   "source": [],
   "metadata": {
    "collapsed": false
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "quantum-env",
   "language": "python",
   "name": "quantum-env"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}