{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "b753ca88",
   "metadata": {},
   "outputs": [],
   "source": [
    "import chess \n",
    "from chess import pgn\n",
    "import os\n",
    "from tqdm import tqdm\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "import numpy as np\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a116056b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 93%|█████████▎| 14/15 [03:50<00:16, 16.44s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Games loaded: 61363\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# hàm để load các file pgn\n",
    "def load_pgn(file_path):\n",
    "    games = []\n",
    "    with open(file_path, 'r') as pgn_file:\n",
    "        while True:\n",
    "            game = pgn.read_game(pgn_file)\n",
    "            if game is None:\n",
    "                break\n",
    "            games.append(game)\n",
    "    return games\n",
    "\n",
    "files = [file for file in os.listdir(\"data\") if file.endswith(\".pgn\")]\n",
    "LIMIT_OF_FILES = len(files) #load all files\n",
    "game = []\n",
    "i = 1\n",
    "for file in tqdm(files):\n",
    "    game.extend(load_pgn(f\"data/{file}\"))\n",
    "    if i >= LIMIT_OF_FILES:\n",
    "        break\n",
    "    i += 1\n",
    "print(\"Games loaded:\", len(game))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "8c876069",
   "metadata": {},
   "outputs": [],
   "source": [
    "from chess import Board\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9ae41f58",
   "metadata": {},
   "outputs": [],
   "source": [
    "def board_to_matrix(board: Board):\n",
    "# Bàn cờ được biểu diễn dưới dạng ma trận 3D với 13 kênh\n",
    "# 12 kênh đầu tiên đại diện cho các loại quân cờ (6 loại cho mỗi màu)\n",
    "# Kênh thứ 13 đại diện cho các nước đi hợp lệ\n",
    "    matrix = np.zeros((13, 8, 8))\n",
    "    piece_map = board.piece_map()\n",
    "    for square, piece in piece_map.items():\n",
    "        row, col = divmod(square, 8)\n",
    "        piece_type = piece.piece_type - 1\n",
    "        piece_color = 0 if piece.color else 6\n",
    "        matrix[piece_type + piece_color, row, col] = 1\n",
    "    legal_moves = board.legal_moves\n",
    "    for move in legal_moves:\n",
    "        to_square = move.to_square\n",
    "        row_to, col_to = divmod(to_square, 8)\n",
    "        matrix[12, row_to, col_to] = 1\n",
    "\n",
    "    return matrix\n",
    "\n",
    "def create_input_for_nn(games): # Tạo dữ liệu đầu vào và nhãn từ các ván cờ\n",
    "    X = []\n",
    "    y = []\n",
    "    for game in games:\n",
    "        board = game.board()\n",
    "        for move in game.mainline_moves():\n",
    "            X.append(board_to_matrix(board))\n",
    "            y.append(move.uci())\n",
    "            board.push(move)\n",
    "    return np.array(X, dtype=np.float32), np.array(y)\n",
    "\n",
    "\n",
    "def encode_moves(moves): # Mã hóa các nước đi thành số nguyên\n",
    "    move_to_int = {move: idx for idx, move in enumerate(set(moves))}\n",
    "    return np.array([move_to_int[move] for move in moves], dtype=np.float32), move_to_int"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5a128d5d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NUM of moves 5538627\n"
     ]
    }
   ],
   "source": [
    "X, y = create_input_for_nn(game)  # Lưu giá trị trả về từ hàm\n",
    "print(\"NUM of moves\", len(y))  # In ra số lượng nước đi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "302f7579",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "y, move_to_int = encode_moves(y) # Mã hóa các nước đi thành số nguyên\n",
    "num_classes = len(move_to_int)\n",
    "\n",
    "X  = torch.tensor(X, dtype = torch.float32) # Chuyển đổi X thành tensor\n",
    "y = torch.tensor(y, dtype = torch.long) # Chuyển đổi y thành tensor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "1602bcf0",
   "metadata": {},
   "outputs": [],
   "source": [
    "from dataset import ChessDataset\n",
    "from model import ChessModel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ca936e40",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using device: cuda\n"
     ]
    }
   ],
   "source": [
    "# Create Dataset and DataLoader\n",
    "from torch.utils.data import DataLoader # type: ignore\n",
    "dataset = ChessDataset(X, y) # Tạo dataset từ tensor X và y\n",
    "dataloader = DataLoader(dataset, batch_size=64, shuffle=True) # Tạo DataLoader để load dữ liệu theo batch size 64\n",
    "\n",
    "# Check for GPU\n",
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "# Model Initialization\n",
    "model = ChessModel(num_classes=num_classes).to(device)\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "optimizer = optim.Adam(model.parameters(), lr=0.0001)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "854f506e",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:26<00:00, 152.83it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 51/101, Loss: 3.2092, Time: 9m26s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:25<00:00, 153.09it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 52/101, Loss: 2.5455, Time: 9m25s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:29<00:00, 152.08it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 53/101, Loss: 2.3894, Time: 9m29s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:29<00:00, 151.96it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 54/101, Loss: 2.2996, Time: 9m29s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:29<00:00, 152.01it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 55/101, Loss: 2.2376, Time: 9m29s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:29<00:00, 151.94it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 56/101, Loss: 2.1911, Time: 9m29s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:29<00:00, 151.87it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 57/101, Loss: 2.1532, Time: 9m29s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:29<00:00, 152.07it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 58/101, Loss: 2.1213, Time: 9m29s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:28<00:00, 152.33it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 59/101, Loss: 2.0945, Time: 9m28s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:28<00:00, 152.15it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 60/101, Loss: 2.0712, Time: 9m28s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:28<00:00, 152.31it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 61/101, Loss: 2.0509, Time: 9m28s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:28<00:00, 152.26it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 62/101, Loss: 2.0328, Time: 9m28s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:27<00:00, 152.39it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 63/101, Loss: 2.0171, Time: 9m27s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:30<00:00, 151.76it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 64/101, Loss: 2.0031, Time: 9m30s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:29<00:00, 151.89it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 65/101, Loss: 1.9905, Time: 9m29s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:32<00:00, 151.06it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 66/101, Loss: 1.9789, Time: 9m32s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:48<00:00, 146.96it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 67/101, Loss: 1.9688, Time: 9m48s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:47<00:00, 147.41it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 68/101, Loss: 1.9591, Time: 9m47s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:41<00:00, 148.76it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 69/101, Loss: 1.9503, Time: 9m41s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:29<00:00, 152.02it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 70/101, Loss: 1.9424, Time: 9m29s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:27<00:00, 152.58it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 71/101, Loss: 1.9349, Time: 9m27s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:28<00:00, 152.16it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 72/101, Loss: 1.9279, Time: 9m28s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:28<00:00, 152.35it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 73/101, Loss: 1.9212, Time: 9m28s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:27<00:00, 152.48it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 74/101, Loss: 1.9154, Time: 9m27s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:28<00:00, 152.35it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 75/101, Loss: 1.9101, Time: 9m28s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:27<00:00, 152.39it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 76/101, Loss: 1.9049, Time: 9m27s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:27<00:00, 152.41it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 77/101, Loss: 1.9002, Time: 9m27s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:27<00:00, 152.44it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 78/101, Loss: 1.8961, Time: 9m27s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:27<00:00, 152.57it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 79/101, Loss: 1.8919, Time: 9m27s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:29<00:00, 151.95it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 80/101, Loss: 1.8881, Time: 9m29s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:20<00:00, 154.40it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 81/101, Loss: 1.8848, Time: 9m20s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:21<00:00, 154.02it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 82/101, Loss: 1.8816, Time: 9m21s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:25<00:00, 152.96it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 83/101, Loss: 1.8788, Time: 9m25s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:28<00:00, 152.32it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 84/101, Loss: 1.8761, Time: 9m28s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:28<00:00, 152.20it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 85/101, Loss: 1.8733, Time: 9m28s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:30<00:00, 151.65it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 86/101, Loss: 1.8709, Time: 9m30s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:26<00:00, 152.75it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 87/101, Loss: 1.8688, Time: 9m26s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:25<00:00, 153.01it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 88/101, Loss: 1.8668, Time: 9m25s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:28<00:00, 152.24it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 89/101, Loss: 1.8647, Time: 9m28s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:26<00:00, 152.85it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 90/101, Loss: 1.8628, Time: 9m26s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:28<00:00, 152.32it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 91/101, Loss: 1.8612, Time: 9m28s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:32<00:00, 151.28it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 92/101, Loss: 1.8597, Time: 9m32s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:31<00:00, 151.42it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 93/101, Loss: 1.8581, Time: 9m31s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:31<00:00, 151.52it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 94/101, Loss: 1.8566, Time: 9m31s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:30<00:00, 151.59it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 95/101, Loss: 1.8553, Time: 9m30s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:31<00:00, 151.53it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 96/101, Loss: 1.8542, Time: 9m31s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:31<00:00, 151.52it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 97/101, Loss: 1.8532, Time: 9m31s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:31<00:00, 151.51it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 98/101, Loss: 1.8519, Time: 9m31s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:31<00:00, 151.56it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 99/101, Loss: 1.8510, Time: 9m31s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 86542/86542 [09:30<00:00, 151.68it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 100/101, Loss: 1.8500, Time: 9m30s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "num_epochs = 50\n",
    "for epoch in range(num_epochs):\n",
    "    start_time = time.time()\n",
    "    model.train()\n",
    "    running_loss = 0.0\n",
    "    for inputs, labels in tqdm(dataloader):\n",
    "        inputs, labels = inputs.to(device), labels.to(device)  # chuyển dữ liệu sang GPU nếu có\n",
    "        optimizer.zero_grad() # trả về gradient về 0\n",
    "\n",
    "        outputs = model(inputs) \n",
    "\n",
    "        # Tính loss và backpropagation\n",
    "        loss = criterion(outputs, labels)\n",
    "        loss.backward()\n",
    "        \n",
    "        # Gradient clipping\n",
    "        torch.nn.utils.clip_grad_norm_(model.parameters(), max_norm=1.0)\n",
    "        \n",
    "        optimizer.step()\n",
    "        running_loss += loss.item()\n",
    "    end_time = time.time()\n",
    "    epoch_time = end_time - start_time\n",
    "    minutes: int = int(epoch_time // 60)\n",
    "    seconds: int = int(epoch_time) - minutes * 60\n",
    "    print(f'Epoch {epoch + 1 + 50}/{num_epochs + 1 + 50}, Loss: {running_loss / len(dataloader):.4f}, Time: {minutes}m{seconds}s')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3521671f",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# lưu model vào models/\n",
    "torch.save(model.state_dict(), \"models/TORCH_1_100EPOCHS.pth\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "c7a53a41",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "\n",
    "with open(\"models/heavy_move_to_int_1\", \"wb\") as file:\n",
    "    pickle.dump(move_to_int, file)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "chess_game",
   "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": "3.13.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}