diff --git a/LICENSE b/LICENSE
new file mode 100644
index 0000000000000000000000000000000000000000..9038d3495e1ebb40dc2f31bddc30937f522ccff6
--- /dev/null
+++ b/LICENSE
@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 ivanovot
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
diff --git a/QuckDrawGAN/__init__.py b/QuckDrawGAN/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..b687a42752e6c7bf1800c1397305eac28684901c
--- /dev/null
+++ b/QuckDrawGAN/__init__.py
@@ -0,0 +1,2 @@
+from .train import train, discriminator_fine_tune
+from .model import Model
diff --git a/QuckDrawGAN/__pycache__/__init__.cpython-312.pyc b/QuckDrawGAN/__pycache__/__init__.cpython-312.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..372c56707e16f707b959ea96a711554eaf621cb8
Binary files /dev/null and b/QuckDrawGAN/__pycache__/__init__.cpython-312.pyc differ
diff --git a/QuckDrawGAN/__pycache__/model.cpython-312.pyc b/QuckDrawGAN/__pycache__/model.cpython-312.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..068e7b3bf0e8d36cc45f1ac443ea8b7948f6080a
Binary files /dev/null and b/QuckDrawGAN/__pycache__/model.cpython-312.pyc differ
diff --git a/QuckDrawGAN/__pycache__/train.cpython-312.pyc b/QuckDrawGAN/__pycache__/train.cpython-312.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..34901e0dab0d0ec951bd6d1ca68b85ff3eda125c
Binary files /dev/null and b/QuckDrawGAN/__pycache__/train.cpython-312.pyc differ
diff --git a/QuckDrawGAN/model.py b/QuckDrawGAN/model.py
new file mode 100644
index 0000000000000000000000000000000000000000..75acd72b33da03f49142c7c19a2695723242ea47
--- /dev/null
+++ b/QuckDrawGAN/model.py
@@ -0,0 +1,89 @@
+import argparse
+import torch
+import numpy as np
+import torchvision.utils as vutils
+from .utils.models import Generator, Discriminator, latent_dim
+import hashlib
+from PIL import Image
+import warnings
+warnings.filterwarnings("ignore", category=RuntimeWarning)
+
+
+class Model:
+    def __init__(self, generator_path, discriminator_path=None):
+        # Определяем устройство для выполнения (GPU или CPU)
+        self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+
+        # Инициализация и загрузка генератора
+        self.generator = Generator(latent_dim).to(self.device)
+        self.generator.load_state_dict(torch.load(generator_path, map_location=self.device, weights_only=True))  # Загружаем веса генератора
+        self.generator.eval()  # Переводим генератор в режим оценки
+
+        # Инициализация дискриминатора, если задан путь к его весам
+        if discriminator_path:
+            self.discriminator = Discriminator().to(self.device)
+            self.discriminator.load_state_dict(torch.load(discriminator_path, map_location=self.device, weights_only=True))  # Загружаем веса дискриминатора
+            self.discriminator.eval()  # Переводим дискриминатор в режим оценки
+        else:
+            self.discriminator = None  # Если дискриминатор не используется
+
+    def generate(self, n=1, seed=None):
+        """Генерирует n изображений. Если дискриминатор загружен, возвращает изображение с наибольшей оценкой дискриминатора."""
+        with torch.no_grad():  # Отключаем градиенты для режима оценки
+            # Установка сида для воспроизводимости, если задан
+            if seed is not None:
+                seed_number = int(hashlib.md5(seed.encode()).hexdigest(), 16) % (2**32)  # Преобразуем текстовый сид в число
+                torch.manual_seed(seed_number)  # Устанавливаем сид для генерации
+
+            # Генерация случайного латентного вектора
+            z = torch.randn(n, latent_dim).to(self.device)
+
+            # Генерация изображений
+            gen_imgs = self.generator(z)
+
+            # Если дискриминатор загружен, выбираем изображение с наилучшей оценкой
+            if self.discriminator:
+                predictions = self.discriminator(gen_imgs).cpu().numpy().flatten()  # Получаем оценки дискриминатора
+                max_pred_idx = predictions.argmax()  # Находим индекс изображения с максимальной оценкой
+                best_img = gen_imgs[max_pred_idx].cpu().squeeze().numpy()  # Преобразуем изображение в формат (H, W)
+                return best_img  # Возвращаем лучшее изображение
+            else:
+                # Если дискриминатор не загружен, возвращаем первое сгенерированное изображение
+                return gen_imgs[0].cpu().squeeze().numpy()  # Преобразуем изображение в формат (H, W)
+
+if __name__ == "__main__":
+    # Определение аргументов командной строки
+    parser = argparse.ArgumentParser(description="Generate image using pretrained GAN model")
+    parser.add_argument('--generator_path', type=str, required=True, help='Path to generator model weights')
+    parser.add_argument('--discriminator_path', type=str, help='Path to discriminator model weights (optional)')
+    parser.add_argument('--output_path', type=str, default='result.png', help='Path to save the generated image')
+    parser.add_argument('--n', type=int, default=1, help='Number of images to generate')
+    parser.add_argument('--seed', type=str, help='Seed for random generation (optional)')
+
+    args = parser.parse_args()
+
+    # Инициализация модели и генерация изображения
+    model = Model(args.generator_path, args.discriminator_path)
+    generated_image = model.generate(n=args.n, seed=args.seed)
+
+    # Нормализация изображения
+    min_val = np.min(generated_image)
+    max_val = np.max(generated_image)
+
+    # Применяем нормализацию
+    normalized_image = (generated_image - min_val) / (max_val - min_val) * 255
+
+    # Приводим к 8-битному формату
+    normalized_image = normalized_image.astype(np.uint8)
+
+    # Проверяем размерность и преобразуем в RGB, если это необходимо
+    if normalized_image.ndim == 2:  # Если изображение в градациях серого
+        # Преобразуем в RGB (64, 64) -> (64, 64, 3)
+        normalized_image = np.stack([normalized_image] * 3, axis=-1)
+    elif normalized_image.shape[2] == 1:  # Если изображение с одним каналом
+        # Удаляем канал и преобразуем в RGB
+        normalized_image = np.squeeze(normalized_image, axis=2)
+
+    # Создаем изображение в формате RGB
+    img = Image.fromarray(normalized_image, mode='RGB')
+    img.save(args.output_path)
diff --git a/QuckDrawGAN/train.py b/QuckDrawGAN/train.py
new file mode 100644
index 0000000000000000000000000000000000000000..21ffc6da5ffe174224645bfd0713d9b24ad2ba37
--- /dev/null
+++ b/QuckDrawGAN/train.py
@@ -0,0 +1,198 @@
+import argparse
+import os
+import logging
+import random
+import torch
+import torch.optim as optim
+import torch.nn as nn
+import torchvision.utils as vutils
+from torch.utils.data import DataLoader
+from tqdm import tqdm
+from .utils.models import Generator, Discriminator, latent_dim
+from .utils.data import DrawDataset
+
+
+def train(epochs, batch_size, data_path, output_path='output', lr_g=0.001, lr_d=0.002, data_max_size=None):
+    # Создание директорий для сохранения изображений и моделей
+    os.makedirs(os.path.join(output_path, 'images'), exist_ok=True)
+    os.makedirs(os.path.join(output_path, 'models'), exist_ok=True)
+    
+    # Инициализация логирования
+    log_file = os.path.join(output_path, 'training_logs.log')
+    with open(log_file, 'w'):
+        pass  # Очищаем файл логов
+    logging.basicConfig(filename=log_file, level=logging.INFO, format='%(asctime)s - %(message)s')
+
+    # Определение устройства для обучения
+    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+    logging.info(f"Using device: {device}")
+
+    logging.info("Loading dataset")
+    dataset = DrawDataset(data_path, data_max_size)
+
+    # Инициализация генератора и дискриминатора
+    generator = Generator(latent_dim).to(device)
+    discriminator = Discriminator().to(device)
+
+    # Оптимизаторы для генератора и дискриминатора
+    optimizer_G = optim.Adam(generator.parameters(), lr=lr_g, betas=(0.9, 0.999))
+    optimizer_D = optim.Adam(discriminator.parameters(), lr=lr_d, betas=(0.9, 0.999))
+
+    # Функция потерь
+    adversarial_loss = nn.L1Loss()
+
+    # Фиксированные векторы шума для генерации изображений в каждом эпохе
+    fix_z = torch.randn(64, latent_dim).to(device)
+
+    # Загрузка данных
+    dataloader = DataLoader(dataset, batch_size=batch_size, shuffle=True, pin_memory=True)
+
+    logging.info("Training started")
+
+    # Основной цикл обучения
+    for epoch in range(epochs):
+        progress_bar = tqdm(dataloader, desc=f"Epoch [{epoch+1}/{epochs}]", bar_format='{l_bar}{bar:12}{r_bar}')
+        generator.train()
+
+        for i, real_imgs in enumerate(progress_bar):
+            real_imgs = real_imgs.to(device)
+            batch_size = real_imgs.size(0)
+
+            # Создание меток для реальных и поддельных изображений
+            valid_labels = torch.full((batch_size, 1), random.uniform(0.7, 1)).to(device)
+            fake_labels = torch.full((batch_size, 1), random.uniform(0, 0.3)).to(device)
+
+            # Обновление дискриминатора
+            optimizer_D.zero_grad()
+
+            # Генерация поддельных изображений
+            z = torch.randn(batch_size, latent_dim).to(device)
+            gen_imgs = generator(z)
+
+            # Вычисление потерь для реальных и поддельных изображений
+            real_preds = discriminator(real_imgs)
+            fake_preds = discriminator(gen_imgs.detach())
+
+            loss_real = adversarial_loss(real_preds, valid_labels)
+            loss_fake = adversarial_loss(fake_preds, fake_labels)
+
+            loss_D = loss_real + loss_fake
+            loss_D.backward()
+            optimizer_D.step()
+
+            # Обновление генератора
+            optimizer_G.zero_grad()
+
+            # Генерация новых поддельных изображений
+            gen_imgs = generator(z)
+
+            # Потери генератора на основе предсказаний дискриминатора
+            fake_preds_for_gen = discriminator(gen_imgs)
+            loss_G = adversarial_loss(fake_preds_for_gen, valid_labels)
+
+            loss_G.backward()
+            optimizer_G.step()
+
+            # Обновление информации в прогресс-баре
+            progress_bar.set_postfix(Loss_D=loss_D.item(), Loss_G=loss_G.item())
+
+        # Логирование итогов эпохи
+        logging.info(f"Epoch [{epoch+1}/{epochs}], Loss_D: {loss_D.item():.4f}, Loss_G: {loss_G.item():.4f}")
+
+        # Сохранение изображений и модели
+        with torch.no_grad():
+            generator.eval()
+            gen_imgs = generator(fix_z)
+            vutils.save_image(gen_imgs.data, os.path.join(output_path, 'images', f'{epoch+1}.png'), nrow=8, normalize=True)
+            torch.save(generator.state_dict(), os.path.join(output_path, 'models', 'generator.pt'))
+            torch.save(discriminator.state_dict(), os.path.join(output_path, 'models', 'discriminator.pt'))
+
+
+def discriminator_fine_tune(generator_file, discriminator_file, data_path, batch_size=64, fine_tune_epochs=10, lr_d=0.002, data_max_size=None):
+    # Определение устройства
+    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+    logging.info(f"Using device: {device}")
+
+    # Загрузка сохранённых моделей
+    generator = Generator(latent_dim).to(device)
+    discriminator = Discriminator().to(device)
+    generator.load_state_dict(torch.load(generator_file, map_location=device, weights_only=True))
+    discriminator.load_state_dict(torch.load(discriminator_file, map_location=device, weights_only=True))
+
+    # Оптимизатор для дискриминатора
+    optimizer_D = optim.Adam(discriminator.parameters(), lr=lr_d, betas=(0.9, 0.999))
+    adversarial_loss = nn.L1Loss()
+
+    # Загрузка данных
+    dataset = DrawDataset(data_path, data_max_size)
+    fine_tune_dataloader = DataLoader(dataset, batch_size=batch_size//2, shuffle=True, pin_memory=True)
+
+    logging.info(f"Fine-tuning discriminator for {fine_tune_epochs} epochs")
+
+    for epoch in range(fine_tune_epochs):
+        progress_bar = tqdm(fine_tune_dataloader, desc=f"Fine-tuning Discriminator [{epoch+1}/{fine_tune_epochs}]", bar_format='{l_bar}{bar:12}{r_bar}')
+        discriminator.train()
+
+        for i, real_imgs in enumerate(progress_bar):
+            real_imgs = real_imgs.to(device)
+            batch_size = real_imgs.size(0)
+
+            # Создание меток для реальных и поддельных изображений
+            valid_labels = torch.full((batch_size, 1), random.uniform(0.7, 1)).to(device)
+            fake_labels = torch.full((batch_size // 2, 1), random.uniform(0, 0.3)).to(device)
+
+            # Обновление дискриминатора
+            optimizer_D.zero_grad()
+
+            # Генерация поддельных изображений
+            z = torch.randn(batch_size // 2, latent_dim).to(device)
+            gen_imgs = generator(z)
+
+            # Потери для реальных и поддельных изображений
+            real_preds = discriminator(real_imgs)
+            fake_preds = discriminator(gen_imgs.detach())
+
+            loss_real = adversarial_loss(real_preds, valid_labels)
+            loss_fake = adversarial_loss(fake_preds, fake_labels)
+
+            loss_D = loss_real + loss_fake
+            loss_D.backward()
+            optimizer_D.step()
+
+            # Обновление информации в прогресс-баре
+            progress_bar.set_postfix(Loss_D=loss_D.item())
+
+        # Логирование результатов дообучения
+        logging.info(f"Fine-tune Epoch [{epoch+1}/{fine_tune_epochs}], Loss_D: {loss_D.item():.4f}")
+
+    # Сохранение обновлённого дискриминатора
+    torch.save(discriminator.state_dict(), os.path.join(os.path.dirname(discriminator_file), 'discriminator_fine_tuned.pt'))
+
+
+# Определение аргументов командной строки
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Train GAN model with specified parameters.")
+    
+    # Основные аргументы для функции train
+    parser.add_argument('--epochs', type=int, default=10, help='Number of epochs for training')
+    parser.add_argument('--batch_size', type=int, default=64, help='Batch size for training')
+    parser.add_argument('--data_path', type=str, required=True, help='Path to training data')
+    parser.add_argument('--output_path', type=str, default='output', help='Directory to save outputs')
+    parser.add_argument('--lr_g', type=float, default=0.001, help='Learning rate for generator')
+    parser.add_argument('--lr_d', type=float, default=0.002, help='Learning rate for discriminator')
+    parser.add_argument('--data_max_size', type=int, default=None, help='Maximum size of data to use')
+
+    # Аргументы для дообучения дискриминатора
+    parser.add_argument('--fine_tune', action='store_true', help='Fine-tune discriminator')
+    parser.add_argument('--generator_file', type=str, help='Path to generator weights for fine-tuning')
+    parser.add_argument('--discriminator_file', type=str, help='Path to discriminator weights for fine-tuning')
+    parser.add_argument('--fine_tune_epochs', type=int, default=10, help='Number of epochs for fine-tuning discriminator')
+    
+    args = parser.parse_args()
+    
+    if args.fine_tune:
+        # Запуск функции дообучения дискриминатора
+        discriminator_fine_tune(args.generator_file, args.discriminator_file, args.data_path, args.batch_size, args.fine_tune_epochs, args.lr_d, args.data_max_size)
+    else:
+        # Запуск основной функции тренировки
+        train(args.epochs, args.batch_size, args.data_path, args.output_path, args.lr_g, args.lr_d, args.data_max_size)
\ No newline at end of file
diff --git a/QuckDrawGAN/utils/__pycache__/data.cpython-312.pyc b/QuckDrawGAN/utils/__pycache__/data.cpython-312.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..ada353e977eb52c935c8887fbf054629d2d7afdb
Binary files /dev/null and b/QuckDrawGAN/utils/__pycache__/data.cpython-312.pyc differ
diff --git a/QuckDrawGAN/utils/__pycache__/models.cpython-312.pyc b/QuckDrawGAN/utils/__pycache__/models.cpython-312.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..d966cbbbc505dcc85904c78bfd80bcc467ccb468
Binary files /dev/null and b/QuckDrawGAN/utils/__pycache__/models.cpython-312.pyc differ
diff --git a/QuckDrawGAN/utils/data.py b/QuckDrawGAN/utils/data.py
new file mode 100644
index 0000000000000000000000000000000000000000..b4198bdf145c846b23231fccc02561bb7a5eebe6
--- /dev/null
+++ b/QuckDrawGAN/utils/data.py
@@ -0,0 +1,63 @@
+import torch
+from torch.utils.data import Dataset
+import pandas as pd
+from PIL import Image, ImageDraw
+import numpy as np
+
+class DrawDataset(Dataset):
+    def __init__(self, file_path, data_max_size=None):
+        # Загрузка данных из файла формата JSON
+        self.data = pd.read_json(file_path, lines=True)
+        # Отбор только распознанных рисунков
+        self.data = self.data[self.data['recognized'] == True]
+        # Ограничение размера набора данных, если указано
+        if data_max_size and len(self.data) > data_max_size:
+            self.data = self.data[:data_max_size]
+        
+        self.images = self.data['drawing'].values
+        self.processed_images = []
+
+        # Преобразование набора в изображения и нормализация
+        for raw_drawing in self.images:
+            img = self.stroke_to_image(raw_drawing)
+            img = np.array(img).astype(np.float32) / 255.0  # Нормализация изображения в диапазон [0, 1]
+            img = torch.from_numpy(img)  # Преобразование в тензор PyTorch
+            self.processed_images.append(img.unsqueeze(0))  # Добавление оси канала (1, 64, 64)
+
+    def stroke_to_image(self, raw_drawing):
+        # Коэффициенты для изменения размера изображения и его улучшения
+        scale_factor = 0.22  # Масштаб для уменьшения координат рисунков
+        upscale_factor = 8  # Коэффициент увеличения для получения плавных линий
+        original_size = 64  # Окончательный размер изображения
+        large_size = original_size * upscale_factor  # Увеличенный размер для рисования линий
+
+        # Преобразование координат линий с масштабированием и смещением
+        polylines = (
+            zip([(x + 25) * scale_factor * upscale_factor for x in polyline[0]],
+                [(y + 25) * scale_factor * upscale_factor for y in polyline[1]])
+            for polyline in raw_drawing if len(polyline) == 2
+        )
+
+        # Преобразуем набор линий в список для последующего рисования
+        polylines_list = [list(polyline) for polyline in polylines]
+        
+        # Создание пустого увеличенного изображения
+        pil_img = Image.new("L", (large_size, large_size), 255)  # Черно-белое изображение, белый фон
+        d = ImageDraw.Draw(pil_img)
+        
+        # Рисование линий с учетом масштабирования и увеличенной толщины
+        for polyline in polylines_list:
+            d.line(polyline, fill=0, width=int(1.5 * upscale_factor))  # Линии черного цвета
+        
+        # Масштабирование изображения обратно до 64x64 с использованием LANCZOS для сглаживания
+        pil_img = pil_img.resize((original_size, original_size), Image.Resampling.LANCZOS)
+        
+        return pil_img
+
+    def __len__(self):
+        # Возвращает количество изображений в наборе данных
+        return len(self.images)
+
+    def __getitem__(self, idx):
+        # Возвращает обработанное изображение по индексу
+        return self.processed_images[idx]
diff --git a/QuckDrawGAN/utils/models.py b/QuckDrawGAN/utils/models.py
new file mode 100644
index 0000000000000000000000000000000000000000..9054a961771d1220c10659cf66ec182aa354a087
--- /dev/null
+++ b/QuckDrawGAN/utils/models.py
@@ -0,0 +1,107 @@
+import torch.nn as nn
+
+latent_dim = 100  # Размерность латентного пространства (входного шума для генератора)
+
+# Класс генератора для генерации изображений из латентного пространства
+class Generator(nn.Module):
+    def __init__(self, latent_dim):
+        super(Generator, self).__init__()
+
+        # Размеры начального изображения и количество каналов для начала транспонированных операций
+        self.init_size = 4  # Размер изображения после первой линейной трансформации
+        self.start_channels = 512  # Количество каналов на первом этапе
+
+        # Последовательная модель генератора
+        self.model = nn.Sequential(
+            # Линейное преобразование латентного вектора в развернутую форму для дальнейшего увеличения
+            nn.Linear(latent_dim, self.start_channels * self.init_size ** 2),
+            nn.BatchNorm1d(self.start_channels * self.init_size ** 2, 0.8),
+            nn.LeakyReLU(0.2, inplace=True),
+
+            # Преобразование в 4D тензор для начала операций с изображениями
+            nn.Unflatten(1, (self.start_channels, self.init_size, self.init_size)),
+
+            # Начало операций с изображением (увеличение размера)
+            nn.Upsample(scale_factor=2),  # Увеличение размера изображения в 2 раза
+
+            # Сверточные слои с уменьшением количества каналов и последующими нелинейностями
+            nn.Conv2d(self.start_channels, self.start_channels // 3, 3, stride=1, padding=1),
+            nn.BatchNorm2d(self.start_channels // 3, 0.8),
+            nn.LeakyReLU(0.2, inplace=True),
+
+            nn.Upsample(scale_factor=2),  # Еще одно увеличение
+
+            nn.Conv2d(self.start_channels // 3, self.start_channels // 4, 3, stride=1, padding=1),
+            nn.BatchNorm2d(self.start_channels // 4, 0.8),
+            nn.LeakyReLU(0.2, inplace=True),
+
+            nn.Upsample(scale_factor=2),  # Третье увеличение
+
+            nn.Conv2d(self.start_channels // 4, self.start_channels // 6, 3, stride=1, padding=1),
+            nn.BatchNorm2d(self.start_channels // 6, 0.8),
+            nn.LeakyReLU(0.2, inplace=True),
+
+            nn.Upsample(scale_factor=2),  # Четвертое увеличение
+
+            nn.Conv2d(self.start_channels // 6, self.start_channels // 8, 3, stride=1, padding=1),
+            nn.BatchNorm2d(self.start_channels // 8, 0.8),
+            nn.LeakyReLU(0.2, inplace=True),
+
+            # Последний сверточный слой для вывода изображения размером 1xWxH с функцией активации Тангенс
+            nn.Conv2d(self.start_channels // 8, 1, 3, stride=1, padding=1),
+            nn.Tanh()  # Приведение значений пикселей в диапазон [-1, 1]
+        )
+
+    def forward(self, z):
+        # Прямое распространение через сеть генератора
+        out = self.model(z)
+        return out  # Возвращаем сгенерированное изображение
+
+
+# Класс дискриминатора для различения реальных и сгенерированных изображений
+class Discriminator(nn.Module):
+    def __init__(self):
+        super(Discriminator, self).__init__()
+
+        # Последовательная модель дискриминатора
+        self.model = nn.Sequential(
+            # Первый сверточный блок
+            nn.Conv2d(1, 64, 3, stride=2, padding=1),  # Уменьшает размер изображения до (64, 32, 32)
+            nn.BatchNorm2d(64, 0.8),
+            nn.LeakyReLU(0.2, inplace=True),
+            nn.Dropout(0.3),  # Вероятность выключения нейронов для регуляризации
+
+            # Второй сверточный блок
+            nn.Conv2d(64, 128, 3, stride=2, padding=1),  # Уменьшает размер до (128, 16, 16)
+            nn.BatchNorm2d(128, 0.8),
+            nn.LeakyReLU(0.2, inplace=True),
+            nn.Dropout(0.3),
+
+            # Третий сверточный блок
+            nn.Conv2d(128, 256, 3, stride=2, padding=1),  # Уменьшает размер до (256, 8, 8)
+            nn.BatchNorm2d(256, 0.8),
+            nn.LeakyReLU(0.2, inplace=True),
+            nn.Dropout(0.3),
+
+            # Четвертый сверточный блок
+            nn.Conv2d(256, 256, 3, stride=1, padding=1),  # Поддерживает размер (256, 8, 8)
+            nn.BatchNorm2d(256, 0.8),
+            nn.LeakyReLU(0.2, inplace=True),
+            nn.MaxPool2d(2),  # Уменьшает размер до (256, 4, 4)
+            nn.Dropout(0.3),
+
+            # Пятый сверточный блок
+            nn.Conv2d(256, 512, 3, stride=1, padding=1),  # Уменьшает размер до (512, 2, 2)
+            nn.BatchNorm2d(512, 0.8),
+            nn.LeakyReLU(0.2, inplace=True),
+            nn.MaxPool2d(2),  # Размер до (512, 1, 1)
+
+            # Преобразование в плоский вектор для классификации
+            nn.Flatten(),  # Преобразует изображение в вектор (512 * 2 * 2 = 2048)
+            nn.Linear(512 * 2 * 2, 1)  # Полносвязный слой для получения одного скалярного выхода
+        )
+
+    def forward(self, img):
+        # Прямое распространение через сеть дискриминатора
+        out = self.model(img)
+        return out  # Возвращаем вероятность принадлежности к классу "реальное" или "сгенерированное"
diff --git a/app.py b/app.py
new file mode 100644
index 0000000000000000000000000000000000000000..7f59e37bd2c34cb003daea03bdf8b32613e67bb6
--- /dev/null
+++ b/app.py
@@ -0,0 +1,71 @@
+import gradio as gr
+import QuckDrawGAN as qd
+import numpy as np
+from PIL import Image
+
+# Загрузим модель
+generator_file = r'pretrained_output/models/generator.pt'
+discriminator_file = r'pretrained_output/models/discriminator_fine_tuned.pt'
+
+# Создаем объект модели
+model = qd.Model(generator_file, discriminator_file)
+
+# Функция для генерации изображения с учетом сида, нормализацией и изменением размера
+def generate_image(n_images=16, seed=""):
+    # Если сид не задан, не передаем его в модель
+    if seed == "":
+        seed = None
+    best_image = model.generate(n_images, seed)  # Генерация с учетом сида
+
+    # Нормализация: находим минимум и максимум в изображении
+    best_image_min = np.min(best_image)
+    best_image_max = np.max(best_image)
+
+    # Нормализуем изображение, чтобы значения были в диапазоне от 0 до 255
+    normalized_image = 255 * (best_image - best_image_min) / (best_image_max - best_image_min)
+
+    # Преобразуем изображение в формат, подходящий для отображения
+    pil_image = Image.fromarray(normalized_image.astype(np.uint8))  # Преобразуем в uint8 для отображения
+    pil_image = pil_image.resize((256, 256), Image.Resampling.LANCZOS)  # Ресайз изображения до 256x256
+
+    return pil_image
+
+# Создаем интерфейс Gradio с помощью Blocks (для большей гибкости)
+with gr.Blocks() as interface:
+    gr.Markdown("# Генератор изображений с использованием QuckDrawGAN")
+    gr.Markdown("Этот интерфейс позволяет генерировать изображения с помощью модели QuckDrawGAN. Настройте количество генерируемых изображений и задайте сид для повторяемости.")
+
+    with gr.Row():
+        # Блок для изображения сверху
+        generated_image = gr.Image(type="pil", label="Сгенерированное изображение", elem_id="generated_image", scale=2)  # Увеличиваем масштаб изображения
+
+    with gr.Row():
+        # Блок параметров и кнопки
+        with gr.Column():
+            seed_input = gr.Textbox(value="", label="Сид (опционально)", interactive=True)
+            num_images = gr.Slider(minimum=1, maximum=1024, value=16, label="Количество изображений для генерации", interactive=True, step=1)
+        
+        # Кнопка генерации изображения справа
+        generate_button = gr.Button("Сгенерировать")
+
+    # Логика для автогенерации при изменении параметров
+    seed_input.change(generate_image, inputs=[num_images, seed_input], outputs=generated_image)
+    num_images.change(generate_image, inputs=[num_images, seed_input], outputs=generated_image)
+
+    # Логика для кнопки генерации
+    generate_button.click(generate_image, inputs=[num_images, seed_input], outputs=generated_image)
+
+    # Автогенерация при старте
+    interface.load(generate_image, inputs=[num_images, seed_input], outputs=generated_image)
+
+    # Стилизация блока изображения (увеличение размера блока)
+    interface.css = """
+    #generated_image {
+        width: 400px;
+        height: 400px;
+        margin-top: 20px;
+    }
+    """
+
+# Запуск интерфейса
+interface.launch()
diff --git a/pretrained_output/images/1.png b/pretrained_output/images/1.png
new file mode 100644
index 0000000000000000000000000000000000000000..a05059e697293f315bc978052e46d35855e6fbeb
Binary files /dev/null and b/pretrained_output/images/1.png differ
diff --git a/pretrained_output/images/10.png b/pretrained_output/images/10.png
new file mode 100644
index 0000000000000000000000000000000000000000..d3ab7a5fc1108e51b160fca4854ceb111c149caa
Binary files /dev/null and b/pretrained_output/images/10.png differ
diff --git a/pretrained_output/images/100.png b/pretrained_output/images/100.png
new file mode 100644
index 0000000000000000000000000000000000000000..2f1a54522a092701a93ae9c584dfa7b65c5afd6a
Binary files /dev/null and b/pretrained_output/images/100.png differ
diff --git a/pretrained_output/images/11.png b/pretrained_output/images/11.png
new file mode 100644
index 0000000000000000000000000000000000000000..6d1bc1036371d07842239f6134ae84ea00364d39
Binary files /dev/null and b/pretrained_output/images/11.png differ
diff --git a/pretrained_output/images/12.png b/pretrained_output/images/12.png
new file mode 100644
index 0000000000000000000000000000000000000000..a37a992f5eaf4ae7da474afd3f3f25d11fbef58a
Binary files /dev/null and b/pretrained_output/images/12.png differ
diff --git a/pretrained_output/images/13.png b/pretrained_output/images/13.png
new file mode 100644
index 0000000000000000000000000000000000000000..5288f3b123b175baf6a3ba4b0a9d28c9d49baca0
Binary files /dev/null and b/pretrained_output/images/13.png differ
diff --git a/pretrained_output/images/14.png b/pretrained_output/images/14.png
new file mode 100644
index 0000000000000000000000000000000000000000..5cc84687c7ea758bd384ff823c47e6e2ef3c7c41
Binary files /dev/null and b/pretrained_output/images/14.png differ
diff --git a/pretrained_output/images/15.png b/pretrained_output/images/15.png
new file mode 100644
index 0000000000000000000000000000000000000000..2cff2d1b8ccfba5f9af614592f6f9ae1a812791b
Binary files /dev/null and b/pretrained_output/images/15.png differ
diff --git a/pretrained_output/images/16.png b/pretrained_output/images/16.png
new file mode 100644
index 0000000000000000000000000000000000000000..a5142cac322698e6558c972a82433ccfa18d73ee
Binary files /dev/null and b/pretrained_output/images/16.png differ
diff --git a/pretrained_output/images/17.png b/pretrained_output/images/17.png
new file mode 100644
index 0000000000000000000000000000000000000000..d8f795d9e02db03292978384b57b3b4f61f33614
Binary files /dev/null and b/pretrained_output/images/17.png differ
diff --git a/pretrained_output/images/18.png b/pretrained_output/images/18.png
new file mode 100644
index 0000000000000000000000000000000000000000..96cb691f4f8030a9a989cccadaef072d63f2cd22
Binary files /dev/null and b/pretrained_output/images/18.png differ
diff --git a/pretrained_output/images/19.png b/pretrained_output/images/19.png
new file mode 100644
index 0000000000000000000000000000000000000000..8e25b2c80916292af6489ad27c87b24f3b6b7754
Binary files /dev/null and b/pretrained_output/images/19.png differ
diff --git a/pretrained_output/images/2.png b/pretrained_output/images/2.png
new file mode 100644
index 0000000000000000000000000000000000000000..b15a6a211eee20cb33e0f03126b81ac9d031388a
Binary files /dev/null and b/pretrained_output/images/2.png differ
diff --git a/pretrained_output/images/20.png b/pretrained_output/images/20.png
new file mode 100644
index 0000000000000000000000000000000000000000..b3e98ce3b011957d54782fd0482ada725d3c5e0d
Binary files /dev/null and b/pretrained_output/images/20.png differ
diff --git a/pretrained_output/images/21.png b/pretrained_output/images/21.png
new file mode 100644
index 0000000000000000000000000000000000000000..d4592f77581352f4be9a31d071fc1ec4fdcdba3a
Binary files /dev/null and b/pretrained_output/images/21.png differ
diff --git a/pretrained_output/images/22.png b/pretrained_output/images/22.png
new file mode 100644
index 0000000000000000000000000000000000000000..9d972850970ed14cadc2066d6ef93ff71837fb8a
Binary files /dev/null and b/pretrained_output/images/22.png differ
diff --git a/pretrained_output/images/23.png b/pretrained_output/images/23.png
new file mode 100644
index 0000000000000000000000000000000000000000..f770cc663a6606be3e657aadff271125ddf0eb24
Binary files /dev/null and b/pretrained_output/images/23.png differ
diff --git a/pretrained_output/images/24.png b/pretrained_output/images/24.png
new file mode 100644
index 0000000000000000000000000000000000000000..21325e555e6d804aa4b9940fb3e85c139f32ca2f
Binary files /dev/null and b/pretrained_output/images/24.png differ
diff --git a/pretrained_output/images/25.png b/pretrained_output/images/25.png
new file mode 100644
index 0000000000000000000000000000000000000000..356504b7953d1de0295af525f53ec968b889716b
Binary files /dev/null and b/pretrained_output/images/25.png differ
diff --git a/pretrained_output/images/26.png b/pretrained_output/images/26.png
new file mode 100644
index 0000000000000000000000000000000000000000..517c3ab6b4bce60b317f9c5363a4b01f5e6b9a83
Binary files /dev/null and b/pretrained_output/images/26.png differ
diff --git a/pretrained_output/images/27.png b/pretrained_output/images/27.png
new file mode 100644
index 0000000000000000000000000000000000000000..6ba3a03a412e52042d6794656f2522a9a4b5c434
Binary files /dev/null and b/pretrained_output/images/27.png differ
diff --git a/pretrained_output/images/28.png b/pretrained_output/images/28.png
new file mode 100644
index 0000000000000000000000000000000000000000..b3910b0c068b5175bb6e62eb182e3d2c18a29c78
Binary files /dev/null and b/pretrained_output/images/28.png differ
diff --git a/pretrained_output/images/29.png b/pretrained_output/images/29.png
new file mode 100644
index 0000000000000000000000000000000000000000..6905ea277579b09ab350dacc477037d65503e6aa
Binary files /dev/null and b/pretrained_output/images/29.png differ
diff --git a/pretrained_output/images/3.png b/pretrained_output/images/3.png
new file mode 100644
index 0000000000000000000000000000000000000000..08d408b16108a1ec9fdc90a1cf7aa3030ecbe96a
Binary files /dev/null and b/pretrained_output/images/3.png differ
diff --git a/pretrained_output/images/30.png b/pretrained_output/images/30.png
new file mode 100644
index 0000000000000000000000000000000000000000..4efa7f4b19a8d4c505baf9cf25d845399c3f7439
Binary files /dev/null and b/pretrained_output/images/30.png differ
diff --git a/pretrained_output/images/31.png b/pretrained_output/images/31.png
new file mode 100644
index 0000000000000000000000000000000000000000..51851668d66a792bed88214abd74fcc887f7d22f
Binary files /dev/null and b/pretrained_output/images/31.png differ
diff --git a/pretrained_output/images/32.png b/pretrained_output/images/32.png
new file mode 100644
index 0000000000000000000000000000000000000000..7ff837f1001b44b3c58416be9a882f38985d5662
Binary files /dev/null and b/pretrained_output/images/32.png differ
diff --git a/pretrained_output/images/33.png b/pretrained_output/images/33.png
new file mode 100644
index 0000000000000000000000000000000000000000..3c9a831dacd00aa500db4097597dcc9f16095743
Binary files /dev/null and b/pretrained_output/images/33.png differ
diff --git a/pretrained_output/images/34.png b/pretrained_output/images/34.png
new file mode 100644
index 0000000000000000000000000000000000000000..54de11df72698f80b51c07706119d1b9ef65baca
Binary files /dev/null and b/pretrained_output/images/34.png differ
diff --git a/pretrained_output/images/35.png b/pretrained_output/images/35.png
new file mode 100644
index 0000000000000000000000000000000000000000..7b90a999d7506bda262c166fe2d07c69c62c08c1
Binary files /dev/null and b/pretrained_output/images/35.png differ
diff --git a/pretrained_output/images/36.png b/pretrained_output/images/36.png
new file mode 100644
index 0000000000000000000000000000000000000000..26c8f58764f8bd136f3275497ce0025f671e0271
Binary files /dev/null and b/pretrained_output/images/36.png differ
diff --git a/pretrained_output/images/37.png b/pretrained_output/images/37.png
new file mode 100644
index 0000000000000000000000000000000000000000..b82a349cd34af5a82bc64c85dd608d4d305d7ddf
Binary files /dev/null and b/pretrained_output/images/37.png differ
diff --git a/pretrained_output/images/38.png b/pretrained_output/images/38.png
new file mode 100644
index 0000000000000000000000000000000000000000..ad3c0a1b6f1696262e29d2098a4fc09da7ace07c
Binary files /dev/null and b/pretrained_output/images/38.png differ
diff --git a/pretrained_output/images/39.png b/pretrained_output/images/39.png
new file mode 100644
index 0000000000000000000000000000000000000000..f443b9efe83ae36400c4ce1b5d4a1fef8658b94f
Binary files /dev/null and b/pretrained_output/images/39.png differ
diff --git a/pretrained_output/images/4.png b/pretrained_output/images/4.png
new file mode 100644
index 0000000000000000000000000000000000000000..1937a7ba0c6258602b0d10a83a63579f39307f7a
Binary files /dev/null and b/pretrained_output/images/4.png differ
diff --git a/pretrained_output/images/40.png b/pretrained_output/images/40.png
new file mode 100644
index 0000000000000000000000000000000000000000..99eba61970fbc988b7d54ff6ca8608e1fae97c2d
Binary files /dev/null and b/pretrained_output/images/40.png differ
diff --git a/pretrained_output/images/41.png b/pretrained_output/images/41.png
new file mode 100644
index 0000000000000000000000000000000000000000..0efbc4e6faa0fd1306d06e8b1b4fecd6621cc71c
Binary files /dev/null and b/pretrained_output/images/41.png differ
diff --git a/pretrained_output/images/42.png b/pretrained_output/images/42.png
new file mode 100644
index 0000000000000000000000000000000000000000..2c909d8f6ca44f04fe26266ba12e37d77a4cd7f4
Binary files /dev/null and b/pretrained_output/images/42.png differ
diff --git a/pretrained_output/images/43.png b/pretrained_output/images/43.png
new file mode 100644
index 0000000000000000000000000000000000000000..99adf84700de1dba8ac98de3cf1281f98f585920
Binary files /dev/null and b/pretrained_output/images/43.png differ
diff --git a/pretrained_output/images/44.png b/pretrained_output/images/44.png
new file mode 100644
index 0000000000000000000000000000000000000000..7df4d1c68439a27f8a14cfa417ba35e986bb204e
Binary files /dev/null and b/pretrained_output/images/44.png differ
diff --git a/pretrained_output/images/45.png b/pretrained_output/images/45.png
new file mode 100644
index 0000000000000000000000000000000000000000..106e1ada26dcbcc92e1bacf8175eda02944268d5
Binary files /dev/null and b/pretrained_output/images/45.png differ
diff --git a/pretrained_output/images/46.png b/pretrained_output/images/46.png
new file mode 100644
index 0000000000000000000000000000000000000000..5ab6dbeee2b0b239dba8c0f21c9c9708890235a0
Binary files /dev/null and b/pretrained_output/images/46.png differ
diff --git a/pretrained_output/images/47.png b/pretrained_output/images/47.png
new file mode 100644
index 0000000000000000000000000000000000000000..2d381a171c319e6110671baf01b3a39620df8246
Binary files /dev/null and b/pretrained_output/images/47.png differ
diff --git a/pretrained_output/images/48.png b/pretrained_output/images/48.png
new file mode 100644
index 0000000000000000000000000000000000000000..237640d828cf6ef690802667133874668f7e4c40
Binary files /dev/null and b/pretrained_output/images/48.png differ
diff --git a/pretrained_output/images/49.png b/pretrained_output/images/49.png
new file mode 100644
index 0000000000000000000000000000000000000000..e87d3848c3041178b9463c85308cac89c7df25fc
Binary files /dev/null and b/pretrained_output/images/49.png differ
diff --git a/pretrained_output/images/5.png b/pretrained_output/images/5.png
new file mode 100644
index 0000000000000000000000000000000000000000..a68ff2cc8b40e090b658cb9857144dbcc6cb8c31
Binary files /dev/null and b/pretrained_output/images/5.png differ
diff --git a/pretrained_output/images/50.png b/pretrained_output/images/50.png
new file mode 100644
index 0000000000000000000000000000000000000000..e22295b2b424a69594fec017e5ea47ba05a4f97c
Binary files /dev/null and b/pretrained_output/images/50.png differ
diff --git a/pretrained_output/images/51.png b/pretrained_output/images/51.png
new file mode 100644
index 0000000000000000000000000000000000000000..cd8e40143db197067905fd7277fb89cd3ce26a68
Binary files /dev/null and b/pretrained_output/images/51.png differ
diff --git a/pretrained_output/images/52.png b/pretrained_output/images/52.png
new file mode 100644
index 0000000000000000000000000000000000000000..64f3bcd5f83f3038cb87cf9f00393ca2cd31f4af
Binary files /dev/null and b/pretrained_output/images/52.png differ
diff --git a/pretrained_output/images/53.png b/pretrained_output/images/53.png
new file mode 100644
index 0000000000000000000000000000000000000000..34947302fc845b26b12c50261b4fee048910073a
Binary files /dev/null and b/pretrained_output/images/53.png differ
diff --git a/pretrained_output/images/54.png b/pretrained_output/images/54.png
new file mode 100644
index 0000000000000000000000000000000000000000..d83c47370bf58df224c883cb3fdac6a0a904fa8d
Binary files /dev/null and b/pretrained_output/images/54.png differ
diff --git a/pretrained_output/images/55.png b/pretrained_output/images/55.png
new file mode 100644
index 0000000000000000000000000000000000000000..7e1f1fc335bae66fe2644e407e664a4f01a39e3c
Binary files /dev/null and b/pretrained_output/images/55.png differ
diff --git a/pretrained_output/images/56.png b/pretrained_output/images/56.png
new file mode 100644
index 0000000000000000000000000000000000000000..cb7622bf7960edccd616dafe7e544f8f3456148f
Binary files /dev/null and b/pretrained_output/images/56.png differ
diff --git a/pretrained_output/images/57.png b/pretrained_output/images/57.png
new file mode 100644
index 0000000000000000000000000000000000000000..577931a69a9c8643e81f38129cab966eda476a3e
Binary files /dev/null and b/pretrained_output/images/57.png differ
diff --git a/pretrained_output/images/58.png b/pretrained_output/images/58.png
new file mode 100644
index 0000000000000000000000000000000000000000..dcb31e0f7c4df74b30de958f1eba358695b97083
Binary files /dev/null and b/pretrained_output/images/58.png differ
diff --git a/pretrained_output/images/59.png b/pretrained_output/images/59.png
new file mode 100644
index 0000000000000000000000000000000000000000..efadfb70d9ea222c84e26847078ac28360c60ad8
Binary files /dev/null and b/pretrained_output/images/59.png differ
diff --git a/pretrained_output/images/6.png b/pretrained_output/images/6.png
new file mode 100644
index 0000000000000000000000000000000000000000..109b79982b47d52ecc2de926d6dfd4a290afea04
Binary files /dev/null and b/pretrained_output/images/6.png differ
diff --git a/pretrained_output/images/60.png b/pretrained_output/images/60.png
new file mode 100644
index 0000000000000000000000000000000000000000..cf894ce5afadc8429d4988462d45dce67f83af44
Binary files /dev/null and b/pretrained_output/images/60.png differ
diff --git a/pretrained_output/images/61.png b/pretrained_output/images/61.png
new file mode 100644
index 0000000000000000000000000000000000000000..953e2dd8950abf25f2d390778e978dccd39a7769
Binary files /dev/null and b/pretrained_output/images/61.png differ
diff --git a/pretrained_output/images/62.png b/pretrained_output/images/62.png
new file mode 100644
index 0000000000000000000000000000000000000000..53ff678f06263a4834be94ca11ca4571d104429f
Binary files /dev/null and b/pretrained_output/images/62.png differ
diff --git a/pretrained_output/images/63.png b/pretrained_output/images/63.png
new file mode 100644
index 0000000000000000000000000000000000000000..818e6c9391652d68e3e734528374b5f0f3b5cea9
Binary files /dev/null and b/pretrained_output/images/63.png differ
diff --git a/pretrained_output/images/64.png b/pretrained_output/images/64.png
new file mode 100644
index 0000000000000000000000000000000000000000..05abef31330f52f63336446dabcebd293bd17f79
Binary files /dev/null and b/pretrained_output/images/64.png differ
diff --git a/pretrained_output/images/65.png b/pretrained_output/images/65.png
new file mode 100644
index 0000000000000000000000000000000000000000..aa5c102add5ec90e4b0c2f42d0198d51996efd28
Binary files /dev/null and b/pretrained_output/images/65.png differ
diff --git a/pretrained_output/images/66.png b/pretrained_output/images/66.png
new file mode 100644
index 0000000000000000000000000000000000000000..d2b22ccb52d1b6375322bbf3b54e92d64fc4214e
Binary files /dev/null and b/pretrained_output/images/66.png differ
diff --git a/pretrained_output/images/67.png b/pretrained_output/images/67.png
new file mode 100644
index 0000000000000000000000000000000000000000..746a8749110bbae67414edd606b032b746b9458e
Binary files /dev/null and b/pretrained_output/images/67.png differ
diff --git a/pretrained_output/images/68.png b/pretrained_output/images/68.png
new file mode 100644
index 0000000000000000000000000000000000000000..aaecfb458a5e804298a09d83a81ee5648cf7731d
Binary files /dev/null and b/pretrained_output/images/68.png differ
diff --git a/pretrained_output/images/69.png b/pretrained_output/images/69.png
new file mode 100644
index 0000000000000000000000000000000000000000..8e9a70f11a61c0ccd7fd5cc1cf6891494bf396a4
Binary files /dev/null and b/pretrained_output/images/69.png differ
diff --git a/pretrained_output/images/7.png b/pretrained_output/images/7.png
new file mode 100644
index 0000000000000000000000000000000000000000..3afc9f7ba968da8525c94a588bbfa13d34ddc8e3
Binary files /dev/null and b/pretrained_output/images/7.png differ
diff --git a/pretrained_output/images/70.png b/pretrained_output/images/70.png
new file mode 100644
index 0000000000000000000000000000000000000000..1e71d8af6f90e4cc06df8d284c60fa324bb8df61
Binary files /dev/null and b/pretrained_output/images/70.png differ
diff --git a/pretrained_output/images/71.png b/pretrained_output/images/71.png
new file mode 100644
index 0000000000000000000000000000000000000000..9897cb1e46ab0dd297f00438f2fd5ce4e657dbe2
Binary files /dev/null and b/pretrained_output/images/71.png differ
diff --git a/pretrained_output/images/72.png b/pretrained_output/images/72.png
new file mode 100644
index 0000000000000000000000000000000000000000..6179907b7f24754871a009cbd962b826520a30de
Binary files /dev/null and b/pretrained_output/images/72.png differ
diff --git a/pretrained_output/images/73.png b/pretrained_output/images/73.png
new file mode 100644
index 0000000000000000000000000000000000000000..0babce739dca4c496b15cff30d3a42781d24c556
Binary files /dev/null and b/pretrained_output/images/73.png differ
diff --git a/pretrained_output/images/74.png b/pretrained_output/images/74.png
new file mode 100644
index 0000000000000000000000000000000000000000..a2973e3eb8d5444a8202f1a0fab8b7f77d3180bb
Binary files /dev/null and b/pretrained_output/images/74.png differ
diff --git a/pretrained_output/images/75.png b/pretrained_output/images/75.png
new file mode 100644
index 0000000000000000000000000000000000000000..285c4d06c5c52a6fc4c6e99344c31056b1486568
Binary files /dev/null and b/pretrained_output/images/75.png differ
diff --git a/pretrained_output/images/76.png b/pretrained_output/images/76.png
new file mode 100644
index 0000000000000000000000000000000000000000..4f63a0647bce01c8b6939958c64cb9d44501ceb7
Binary files /dev/null and b/pretrained_output/images/76.png differ
diff --git a/pretrained_output/images/77.png b/pretrained_output/images/77.png
new file mode 100644
index 0000000000000000000000000000000000000000..a28df03a81c374420f24ad4c7d3d6a06a58de777
Binary files /dev/null and b/pretrained_output/images/77.png differ
diff --git a/pretrained_output/images/78.png b/pretrained_output/images/78.png
new file mode 100644
index 0000000000000000000000000000000000000000..4e0736b2bece813796f3a8f7a7f97f4d4c75e718
Binary files /dev/null and b/pretrained_output/images/78.png differ
diff --git a/pretrained_output/images/79.png b/pretrained_output/images/79.png
new file mode 100644
index 0000000000000000000000000000000000000000..4af28f10d7d7e3202bda2b05df9fd7dc66a8726b
Binary files /dev/null and b/pretrained_output/images/79.png differ
diff --git a/pretrained_output/images/8.png b/pretrained_output/images/8.png
new file mode 100644
index 0000000000000000000000000000000000000000..941ec264029a979e34f92250451239b39a5eca85
Binary files /dev/null and b/pretrained_output/images/8.png differ
diff --git a/pretrained_output/images/80.png b/pretrained_output/images/80.png
new file mode 100644
index 0000000000000000000000000000000000000000..ac0c2cb396f3b96b7083c03ea497399e05aa10cc
Binary files /dev/null and b/pretrained_output/images/80.png differ
diff --git a/pretrained_output/images/81.png b/pretrained_output/images/81.png
new file mode 100644
index 0000000000000000000000000000000000000000..f30742af742414391d16cd287494569d3aaf6819
Binary files /dev/null and b/pretrained_output/images/81.png differ
diff --git a/pretrained_output/images/82.png b/pretrained_output/images/82.png
new file mode 100644
index 0000000000000000000000000000000000000000..e3d8a5d72a59594865c33c19b2216dbbeb2e4d2e
Binary files /dev/null and b/pretrained_output/images/82.png differ
diff --git a/pretrained_output/images/83.png b/pretrained_output/images/83.png
new file mode 100644
index 0000000000000000000000000000000000000000..1f4b3b6036bd2a418a145a0fe38a6e368994cfab
Binary files /dev/null and b/pretrained_output/images/83.png differ
diff --git a/pretrained_output/images/84.png b/pretrained_output/images/84.png
new file mode 100644
index 0000000000000000000000000000000000000000..b7400115e7d317dc3f6a97d8147f18a7bd1a731e
Binary files /dev/null and b/pretrained_output/images/84.png differ
diff --git a/pretrained_output/images/85.png b/pretrained_output/images/85.png
new file mode 100644
index 0000000000000000000000000000000000000000..5de38020a739fd9197b007bbaf3f379cb2ed49ab
Binary files /dev/null and b/pretrained_output/images/85.png differ
diff --git a/pretrained_output/images/86.png b/pretrained_output/images/86.png
new file mode 100644
index 0000000000000000000000000000000000000000..2cd6c5dd4c032c873a4b43c91f2f5a37c69d8604
Binary files /dev/null and b/pretrained_output/images/86.png differ
diff --git a/pretrained_output/images/87.png b/pretrained_output/images/87.png
new file mode 100644
index 0000000000000000000000000000000000000000..26c4834fa694e6b527f6fabe41fb1b5b6374ee87
Binary files /dev/null and b/pretrained_output/images/87.png differ
diff --git a/pretrained_output/images/88.png b/pretrained_output/images/88.png
new file mode 100644
index 0000000000000000000000000000000000000000..bbe87bac6a65f066d77375199d6eba0d29f3de9f
Binary files /dev/null and b/pretrained_output/images/88.png differ
diff --git a/pretrained_output/images/89.png b/pretrained_output/images/89.png
new file mode 100644
index 0000000000000000000000000000000000000000..c19b73bfb663eaa2374bc211aaac3b6a88a144a0
Binary files /dev/null and b/pretrained_output/images/89.png differ
diff --git a/pretrained_output/images/9.png b/pretrained_output/images/9.png
new file mode 100644
index 0000000000000000000000000000000000000000..b529ad916efa37e077a623eca694ee299644c3ef
Binary files /dev/null and b/pretrained_output/images/9.png differ
diff --git a/pretrained_output/images/90.png b/pretrained_output/images/90.png
new file mode 100644
index 0000000000000000000000000000000000000000..edf908a1ff815ded3b9ea58b9a823da950aa20d9
Binary files /dev/null and b/pretrained_output/images/90.png differ
diff --git a/pretrained_output/images/91.png b/pretrained_output/images/91.png
new file mode 100644
index 0000000000000000000000000000000000000000..da03b2932c0e3dbe033f5e71555ee4005d5ede5a
Binary files /dev/null and b/pretrained_output/images/91.png differ
diff --git a/pretrained_output/images/92.png b/pretrained_output/images/92.png
new file mode 100644
index 0000000000000000000000000000000000000000..89e86eef7d99adb43685bd4d6e3ee4d25c96bdb6
Binary files /dev/null and b/pretrained_output/images/92.png differ
diff --git a/pretrained_output/images/93.png b/pretrained_output/images/93.png
new file mode 100644
index 0000000000000000000000000000000000000000..c15509e033d75812de2f38be0adef9ca452d7d63
Binary files /dev/null and b/pretrained_output/images/93.png differ
diff --git a/pretrained_output/images/94.png b/pretrained_output/images/94.png
new file mode 100644
index 0000000000000000000000000000000000000000..d1b5f3541d205bd17a28bfb1ffabbb9c063f4105
Binary files /dev/null and b/pretrained_output/images/94.png differ
diff --git a/pretrained_output/images/95.png b/pretrained_output/images/95.png
new file mode 100644
index 0000000000000000000000000000000000000000..02fa050944169454244d7329410ec3e384021422
Binary files /dev/null and b/pretrained_output/images/95.png differ
diff --git a/pretrained_output/images/96.png b/pretrained_output/images/96.png
new file mode 100644
index 0000000000000000000000000000000000000000..d8ff37ee9648468da31e6f85d5fa472e1d262e16
Binary files /dev/null and b/pretrained_output/images/96.png differ
diff --git a/pretrained_output/images/97.png b/pretrained_output/images/97.png
new file mode 100644
index 0000000000000000000000000000000000000000..121c0099cf9619c78a9b868147b12ffcecab73a4
Binary files /dev/null and b/pretrained_output/images/97.png differ
diff --git a/pretrained_output/images/98.png b/pretrained_output/images/98.png
new file mode 100644
index 0000000000000000000000000000000000000000..ba4e69d75b563b92d1270ca628a93e557cefcde5
Binary files /dev/null and b/pretrained_output/images/98.png differ
diff --git a/pretrained_output/images/99.png b/pretrained_output/images/99.png
new file mode 100644
index 0000000000000000000000000000000000000000..a12ce5d34abc63e3df4e14717cc4a0cd1ccdd1b5
Binary files /dev/null and b/pretrained_output/images/99.png differ
diff --git a/pretrained_output/models/discriminator.pt b/pretrained_output/models/discriminator.pt
new file mode 100644
index 0000000000000000000000000000000000000000..6feabce3821391d50d5455bc588a136242cdb293
--- /dev/null
+++ b/pretrained_output/models/discriminator.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d2a42be9a668e28e4fddcfecb477c586f79d5fc387e36336b8499ca70036efb
+size 8599709
diff --git a/pretrained_output/models/discriminator_fine_tuned.pt b/pretrained_output/models/discriminator_fine_tuned.pt
new file mode 100644
index 0000000000000000000000000000000000000000..58a5e72bd833290863a948845c62e2ffadd18b8d
--- /dev/null
+++ b/pretrained_output/models/discriminator_fine_tuned.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb8e9e350d621e3ea7b54421c65a37914fc62cb7436b45904f2edf3a7c3c9465
+size 8600370
diff --git a/pretrained_output/models/generator.pt b/pretrained_output/models/generator.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f6d8657a2cc11d5c7282bee7c5c2f8a5f8d0c813
--- /dev/null
+++ b/pretrained_output/models/generator.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0651743016c3b0b906a187427da4df2c3728f0357f43923d24d5b620393f71d0
+size 7966467
diff --git a/pretrained_output/training_logs.log b/pretrained_output/training_logs.log
new file mode 100644
index 0000000000000000000000000000000000000000..eaddf254de66d8fa6cac9defd344097ffb553ea5
--- /dev/null
+++ b/pretrained_output/training_logs.log
@@ -0,0 +1,103 @@
+2024-10-22 20:46:18,783 - Lodadind data
+2024-10-22 20:46:41,624 - Using device: cuda
+2024-10-22 20:46:41,710 - Start train
+2024-10-22 20:47:12,939 - Epoch [1/100], Batch [469/469], Loss_D: 0.2357, Loss_G: 0.6456
+2024-10-22 20:47:44,361 - Epoch [2/100], Batch [469/469], Loss_D: 0.0845, Loss_G: 0.8505
+2024-10-22 20:48:15,698 - Epoch [3/100], Batch [469/469], Loss_D: 0.5342, Loss_G: 0.4654
+2024-10-22 20:48:47,414 - Epoch [4/100], Batch [469/469], Loss_D: 0.1407, Loss_G: 0.6517
+2024-10-22 20:49:19,053 - Epoch [5/100], Batch [469/469], Loss_D: 0.2737, Loss_G: 0.6624
+2024-10-22 20:49:50,159 - Epoch [6/100], Batch [469/469], Loss_D: 0.4493, Loss_G: 0.5388
+2024-10-22 20:50:20,524 - Epoch [7/100], Batch [469/469], Loss_D: 0.3012, Loss_G: 0.5224
+2024-10-22 20:50:50,792 - Epoch [8/100], Batch [469/469], Loss_D: 0.2237, Loss_G: 0.5728
+2024-10-22 20:51:21,609 - Epoch [9/100], Batch [469/469], Loss_D: 0.4069, Loss_G: 0.8186
+2024-10-22 20:51:51,940 - Epoch [10/100], Batch [469/469], Loss_D: 0.2774, Loss_G: 0.8371
+2024-10-22 20:52:22,319 - Epoch [11/100], Batch [469/469], Loss_D: 0.3095, Loss_G: 0.6746
+2024-10-22 20:52:53,089 - Epoch [12/100], Batch [469/469], Loss_D: 0.3624, Loss_G: 0.6893
+2024-10-22 20:53:23,667 - Epoch [13/100], Batch [469/469], Loss_D: 0.0555, Loss_G: 0.5063
+2024-10-22 20:53:53,992 - Epoch [14/100], Batch [469/469], Loss_D: 0.3214, Loss_G: 0.6874
+2024-10-22 20:54:24,337 - Epoch [15/100], Batch [469/469], Loss_D: 0.3996, Loss_G: 0.7867
+2024-10-22 20:54:54,678 - Epoch [16/100], Batch [469/469], Loss_D: 0.3704, Loss_G: 0.7810
+2024-10-22 20:55:25,018 - Epoch [17/100], Batch [469/469], Loss_D: 0.2363, Loss_G: 0.7805
+2024-10-22 20:55:55,853 - Epoch [18/100], Batch [469/469], Loss_D: 0.1914, Loss_G: 0.7242
+2024-10-22 20:56:26,163 - Epoch [19/100], Batch [469/469], Loss_D: 0.2174, Loss_G: 0.5945
+2024-10-22 20:56:56,776 - Epoch [20/100], Batch [469/469], Loss_D: 0.1038, Loss_G: 0.6807
+2024-10-22 20:57:27,420 - Epoch [21/100], Batch [469/469], Loss_D: 0.1360, Loss_G: 0.6402
+2024-10-22 20:57:58,045 - Epoch [22/100], Batch [469/469], Loss_D: 0.2149, Loss_G: 0.6247
+2024-10-22 20:58:28,345 - Epoch [23/100], Batch [469/469], Loss_D: 0.2966, Loss_G: 0.7582
+2024-10-22 20:58:59,570 - Epoch [24/100], Batch [469/469], Loss_D: 0.4792, Loss_G: 0.4935
+2024-10-22 20:59:30,106 - Epoch [25/100], Batch [469/469], Loss_D: 0.1862, Loss_G: 0.5280
+2024-10-22 21:00:01,345 - Epoch [26/100], Batch [469/469], Loss_D: 0.2429, Loss_G: 0.7221
+2024-10-22 21:00:32,596 - Epoch [27/100], Batch [469/469], Loss_D: 0.1608, Loss_G: 0.6689
+2024-10-22 21:01:02,907 - Epoch [28/100], Batch [469/469], Loss_D: 0.3707, Loss_G: 0.6196
+2024-10-22 21:01:33,277 - Epoch [29/100], Batch [469/469], Loss_D: 0.3974, Loss_G: 0.7429
+2024-10-22 21:02:03,728 - Epoch [30/100], Batch [469/469], Loss_D: 0.1752, Loss_G: 0.5156
+2024-10-22 21:02:34,281 - Epoch [31/100], Batch [469/469], Loss_D: 0.1800, Loss_G: 0.5937
+2024-10-22 21:03:04,891 - Epoch [32/100], Batch [469/469], Loss_D: 0.2021, Loss_G: 0.5945
+2024-10-22 21:03:36,546 - Epoch [33/100], Batch [469/469], Loss_D: 0.1095, Loss_G: 0.8008
+2024-10-22 21:04:06,482 - Epoch [34/100], Batch [469/469], Loss_D: 0.2057, Loss_G: 0.6817
+2024-10-22 21:04:37,275 - Epoch [35/100], Batch [469/469], Loss_D: 0.3957, Loss_G: 0.8120
+2024-10-22 21:05:09,022 - Epoch [36/100], Batch [469/469], Loss_D: 0.2212, Loss_G: 0.5342
+2024-10-22 21:05:41,793 - Epoch [37/100], Batch [469/469], Loss_D: 0.1708, Loss_G: 0.7387
+2024-10-22 21:06:13,644 - Epoch [38/100], Batch [469/469], Loss_D: 0.2813, Loss_G: 0.5862
+2024-10-22 21:06:44,255 - Epoch [39/100], Batch [469/469], Loss_D: 0.3603, Loss_G: 0.5560
+2024-10-22 21:07:14,763 - Epoch [40/100], Batch [469/469], Loss_D: 0.1354, Loss_G: 0.5184
+2024-10-22 21:07:45,404 - Epoch [41/100], Batch [469/469], Loss_D: 0.3317, Loss_G: 0.7580
+2024-10-22 21:08:16,153 - Epoch [42/100], Batch [469/469], Loss_D: 0.4953, Loss_G: 0.3403
+2024-10-22 21:08:46,499 - Epoch [43/100], Batch [469/469], Loss_D: 0.1746, Loss_G: 0.5162
+2024-10-22 21:09:16,777 - Epoch [44/100], Batch [469/469], Loss_D: 0.4298, Loss_G: 0.7153
+2024-10-22 21:09:47,954 - Epoch [45/100], Batch [469/469], Loss_D: 0.1553, Loss_G: 0.5628
+2024-10-22 21:10:18,361 - Epoch [46/100], Batch [469/469], Loss_D: 0.2322, Loss_G: 0.7577
+2024-10-22 21:10:48,607 - Epoch [47/100], Batch [469/469], Loss_D: 0.4900, Loss_G: 0.2370
+2024-10-22 21:11:19,368 - Epoch [48/100], Batch [469/469], Loss_D: 0.2106, Loss_G: 0.6747
+2024-10-22 21:11:50,326 - Epoch [49/100], Batch [469/469], Loss_D: 0.3368, Loss_G: 0.8195
+2024-10-22 21:12:21,397 - Epoch [50/100], Batch [469/469], Loss_D: 0.2430, Loss_G: 0.6201
+2024-10-22 21:12:53,127 - Epoch [51/100], Batch [469/469], Loss_D: 0.1312, Loss_G: 0.5016
+2024-10-22 21:13:24,509 - Epoch [52/100], Batch [469/469], Loss_D: 0.2923, Loss_G: 0.8598
+2024-10-22 21:13:54,773 - Epoch [53/100], Batch [469/469], Loss_D: 0.4443, Loss_G: 0.3379
+2024-10-22 21:14:24,376 - Epoch [54/100], Batch [469/469], Loss_D: 0.0724, Loss_G: 0.5998
+2024-10-22 21:14:53,755 - Epoch [55/100], Batch [469/469], Loss_D: 0.4116, Loss_G: 0.4523
+2024-10-22 21:15:23,129 - Epoch [56/100], Batch [469/469], Loss_D: 0.1418, Loss_G: 0.6783
+2024-10-22 21:15:52,513 - Epoch [57/100], Batch [469/469], Loss_D: 0.2035, Loss_G: 0.5773
+2024-10-22 21:16:21,877 - Epoch [58/100], Batch [469/469], Loss_D: 0.0987, Loss_G: 0.7975
+2024-10-22 21:16:51,250 - Epoch [59/100], Batch [469/469], Loss_D: 0.3031, Loss_G: 0.7860
+2024-10-22 21:17:20,629 - Epoch [60/100], Batch [469/469], Loss_D: 0.2889, Loss_G: 0.6679
+2024-10-22 21:17:50,003 - Epoch [61/100], Batch [469/469], Loss_D: 0.2187, Loss_G: 0.5018
+2024-10-22 21:18:19,373 - Epoch [62/100], Batch [469/469], Loss_D: 0.1392, Loss_G: 0.4457
+2024-10-22 21:18:48,746 - Epoch [63/100], Batch [469/469], Loss_D: 0.2681, Loss_G: 0.8638
+2024-10-22 21:19:19,150 - Epoch [64/100], Batch [469/469], Loss_D: 0.2004, Loss_G: 0.7712
+2024-10-22 21:19:49,767 - Epoch [65/100], Batch [469/469], Loss_D: 0.3238, Loss_G: 0.5592
+2024-10-22 21:20:19,145 - Epoch [66/100], Batch [469/469], Loss_D: 0.2861, Loss_G: 0.8833
+2024-10-22 21:20:48,512 - Epoch [67/100], Batch [469/469], Loss_D: 0.2191, Loss_G: 0.5182
+2024-10-22 21:21:17,874 - Epoch [68/100], Batch [469/469], Loss_D: 0.2530, Loss_G: 0.6281
+2024-10-22 21:21:47,236 - Epoch [69/100], Batch [469/469], Loss_D: 0.4905, Loss_G: 0.6754
+2024-10-22 21:22:16,593 - Epoch [70/100], Batch [469/469], Loss_D: 0.3803, Loss_G: 0.6533
+2024-10-22 21:22:45,953 - Epoch [71/100], Batch [469/469], Loss_D: 0.0733, Loss_G: 0.5627
+2024-10-22 21:23:15,355 - Epoch [72/100], Batch [469/469], Loss_D: 0.1307, Loss_G: 0.6862
+2024-10-22 21:23:44,717 - Epoch [73/100], Batch [469/469], Loss_D: 0.1982, Loss_G: 0.6108
+2024-10-22 21:24:14,095 - Epoch [74/100], Batch [469/469], Loss_D: 0.1413, Loss_G: 0.5810
+2024-10-22 21:24:43,463 - Epoch [75/100], Batch [469/469], Loss_D: 0.2061, Loss_G: 0.6669
+2024-10-22 21:25:14,445 - Epoch [76/100], Batch [469/469], Loss_D: 0.1874, Loss_G: 0.6704
+2024-10-22 21:25:44,807 - Epoch [77/100], Batch [469/469], Loss_D: 0.1832, Loss_G: 0.5598
+2024-10-22 21:26:14,187 - Epoch [78/100], Batch [469/469], Loss_D: 0.2969, Loss_G: 0.7973
+2024-10-22 21:26:43,547 - Epoch [79/100], Batch [469/469], Loss_D: 0.3312, Loss_G: 0.5341
+2024-10-22 21:27:12,955 - Epoch [80/100], Batch [469/469], Loss_D: 0.1982, Loss_G: 0.8777
+2024-10-22 21:27:42,320 - Epoch [81/100], Batch [469/469], Loss_D: 0.3367, Loss_G: 0.7962
+2024-10-22 21:28:11,679 - Epoch [82/100], Batch [469/469], Loss_D: 0.1327, Loss_G: 0.6267
+2024-10-22 21:28:41,056 - Epoch [83/100], Batch [469/469], Loss_D: 0.1279, Loss_G: 0.6073
+2024-10-22 21:29:10,443 - Epoch [84/100], Batch [469/469], Loss_D: 0.2912, Loss_G: 0.6855
+2024-10-22 21:29:39,812 - Epoch [85/100], Batch [469/469], Loss_D: 0.3438, Loss_G: 0.7762
+2024-10-22 21:30:09,185 - Epoch [86/100], Batch [469/469], Loss_D: 0.1983, Loss_G: 0.4470
+2024-10-22 21:30:38,956 - Epoch [87/100], Batch [469/469], Loss_D: 0.2298, Loss_G: 0.8061
+2024-10-22 21:31:09,175 - Epoch [88/100], Batch [469/469], Loss_D: 0.3434, Loss_G: 0.6383
+2024-10-22 21:31:39,594 - Epoch [89/100], Batch [469/469], Loss_D: 0.1607, Loss_G: 0.6672
+2024-10-22 21:32:08,987 - Epoch [90/100], Batch [469/469], Loss_D: 0.3369, Loss_G: 0.4305
+2024-10-22 21:32:38,374 - Epoch [91/100], Batch [469/469], Loss_D: 0.0762, Loss_G: 0.6941
+2024-10-22 21:33:07,753 - Epoch [92/100], Batch [469/469], Loss_D: 0.5284, Loss_G: 0.8560
+2024-10-22 21:33:37,135 - Epoch [93/100], Batch [469/469], Loss_D: 0.3745, Loss_G: 0.5909
+2024-10-22 21:34:06,513 - Epoch [94/100], Batch [469/469], Loss_D: 0.0671, Loss_G: 0.6384
+2024-10-22 21:34:35,913 - Epoch [95/100], Batch [469/469], Loss_D: 0.1428, Loss_G: 0.7203
+2024-10-22 21:35:05,290 - Epoch [96/100], Batch [469/469], Loss_D: 0.2742, Loss_G: 0.6134
+2024-10-22 21:35:34,693 - Epoch [97/100], Batch [469/469], Loss_D: 0.1224, Loss_G: 0.7076
+2024-10-22 21:36:04,086 - Epoch [98/100], Batch [469/469], Loss_D: 0.3681, Loss_G: 0.5916
+2024-10-22 21:36:34,006 - Epoch [99/100], Batch [469/469], Loss_D: 0.5593, Loss_G: 0.8537
+2024-10-22 21:37:04,420 - Epoch [100/100], Batch [469/469], Loss_D: 0.0893, Loss_G: 0.6289
diff --git a/requirements.txt b/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..210b23337b8b496e033f458010898ba6d4a00f25
Binary files /dev/null and b/requirements.txt differ