{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.8.17","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"tpu1vmV38","dataSources":[{"sourceId":21755,"databundleVersionId":1475600,"sourceType":"competition"}],"dockerImageVersionId":30512,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":false}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"%pip install imageio\n%pip install keras-visualizer","metadata":{"execution":{"iopub.status.busy":"2023-06-27T00:27:44.227426Z","iopub.execute_input":"2023-06-27T00:27:44.227831Z","iopub.status.idle":"2023-06-27T00:27:55.268706Z","shell.execute_reply.started":"2023-06-27T00:27:44.227799Z","shell.execute_reply":"2023-06-27T00:27:55.267350Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import os, random, json, PIL, shutil, re, imageio, glob\nimport numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\nfrom kaggle_datasets import KaggleDatasets\nfrom PIL import ImageDraw, ImageFont\nimport tensorflow as tf\nimport tensorflow.keras.layers as L\nimport tensorflow_addons as tfa\nfrom tensorflow.keras import Model, models, losses, optimizers\nfrom tensorflow.keras.callbacks import Callback\nfrom keras_visualizer import visualizer","metadata":{"execution":{"iopub.status.busy":"2023-06-27T00:27:55.270997Z","iopub.execute_input":"2023-06-27T00:27:55.271409Z","iopub.status.idle":"2023-06-27T00:28:35.969843Z","shell.execute_reply.started":"2023-06-27T00:27:55.271365Z","shell.execute_reply":"2023-06-27T00:28:35.968932Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"try:\n    tpu = tf.distribute.cluster_resolver.TPUClusterResolver()\n    print(f'Running on TPU {tpu.master()}')\nexcept ValueError:\n    tpu = None\n\nif tpu:\n    tf.config.experimental_connect_to_cluster(tpu)\n    tf.tpu.experimental.initialize_tpu_system(tpu)\n    strategy = tf.distribute.TPUStrategy(tpu)\nelse:\n    strategy = tf.distribute.get_strategy()\n\n\nREPLICAS = strategy.num_replicas_in_sync\nprint(f'REPLICAS: {REPLICAS}')\nAUTO = tf.data.experimental.AUTOTUNE","metadata":{"execution":{"iopub.status.busy":"2023-06-27T00:28:36.011077Z","iopub.execute_input":"2023-06-27T00:28:36.011375Z","iopub.status.idle":"2023-06-27T00:28:44.338035Z","shell.execute_reply.started":"2023-06-27T00:28:36.011348Z","shell.execute_reply":"2023-06-27T00:28:44.337103Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"HEIGHT = 256\nWIDTH = 256\nCHANNELS = 3\nEPOCHS = 80\nBATCH_SIZE = 4","metadata":{"execution":{"iopub.status.busy":"2023-06-27T00:28:44.339914Z","iopub.execute_input":"2023-06-27T00:28:44.340219Z","iopub.status.idle":"2023-06-27T00:28:44.344549Z","shell.execute_reply.started":"2023-06-27T00:28:44.340192Z","shell.execute_reply":"2023-06-27T00:28:44.343695Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"GCS_PATH = KaggleDatasets().get_gcs_path('gan-getting-started')\n\nMONET_FILENAMES = tf.io.gfile.glob(str(GCS_PATH + '/monet_tfrec/*.tfrec'))\nPHOTO_FILENAMES = tf.io.gfile.glob(str(GCS_PATH + '/photo_tfrec/*.tfrec'))\n\ndef count_data_items(filenames):\n    n = [int(re.compile(r\"-([0-9]*)\\.\").search(filename).group(1)) for filename in filenames]\n    return np.sum(n)\n\nn_monet_samples = count_data_items(MONET_FILENAMES)\nn_photo_samples = count_data_items(PHOTO_FILENAMES)\n\nprint(f'Monet TFRecord files: {len(MONET_FILENAMES)}')\nprint(f'Monet image files: {n_monet_samples}')\nprint(f'Photo TFRecord files: {len(PHOTO_FILENAMES)}')\nprint(f'Photo image files: {n_photo_samples}')","metadata":{"execution":{"iopub.status.busy":"2023-06-27T00:28:44.345507Z","iopub.execute_input":"2023-06-27T00:28:44.345841Z","iopub.status.idle":"2023-06-27T00:28:44.373666Z","shell.execute_reply.started":"2023-06-27T00:28:44.345816Z","shell.execute_reply":"2023-06-27T00:28:44.372786Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":" !fc-list","metadata":{"execution":{"iopub.status.busy":"2023-06-26T20:15:05.167012Z","iopub.execute_input":"2023-06-26T20:15:05.168085Z","iopub.status.idle":"2023-06-26T20:15:09.169917Z","shell.execute_reply.started":"2023-06-26T20:15:05.168037Z","shell.execute_reply":"2023-06-26T20:15:09.168184Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def decode_image(image):\n    image = tf.image.decode_jpeg(image, channels=CHANNELS)\n    image = (tf.cast(image, tf.float32) / 127.5) - 1\n    image = tf.reshape(image, [HEIGHT, WIDTH, CHANNELS])\n    return image\n\ndef read_tfrecord(example):\n    tfrecord_format = {\n        'image_name': tf.io.FixedLenFeature([], tf.string),\n        'image':      tf.io.FixedLenFeature([], tf.string),\n        'target':     tf.io.FixedLenFeature([], tf.string)\n    }\n    example = tf.io.parse_single_example(example, tfrecord_format)\n    image = decode_image(example['image'])\n    return image\n\ndef load_dataset(filenames):\n    dataset = tf.data.TFRecordDataset(filenames)\n    dataset = dataset.map(read_tfrecord, num_parallel_calls=AUTO)\n    return dataset\n\ndef normalize_img(img):\n    img = tf.cast(img, dtype=tf.float32)\n    return (img / 127.5) - 1.0\n\ndef get_dataset(filenames, augment=None, repeat=True, shuffle=True, batch_size=1):\n    dataset = load_dataset(filenames)\n\n    if augment:\n        dataset = dataset.map(augment, num_parallel_calls=AUTO)\n        dataset = dataset.map(lambda x: (augment(x)), \n                              num_parallel_calls=AUTO)\n    if repeat:\n        dataset = dataset.repeat()\n    if shuffle:\n        dataset = dataset.shuffle(512)\n        \n    dataset = dataset.batch(batch_size, drop_remainder=True)\n    dataset = dataset.cache()\n    dataset = dataset.prefetch(AUTO)\n    \n    return dataset\n\ndef display_samples(ds, row, col):\n    ds_iter = iter(ds)\n    plt.figure(figsize=(15, int(15*row/col)))\n    for j in range(row*col):\n        example_sample = next(ds_iter)\n        plt.subplot(row,col,j+1)\n        plt.axis('off')\n        plt.imshow(example_sample[0] * 0.5 + 0.5)\n    plt.show()\n        \ndef display_generated_samples(ds, model, n_samples):\n    ds_iter = iter(ds)\n    for n_sample in range(n_samples):\n        example_sample = next(ds_iter)\n        generated_sample = model.predict(example_sample)\n        \n        plt.subplot(121)\n        plt.title(\"input image\")\n        plt.imshow(example_sample[0] * 0.5 + 0.5)\n        plt.axis('off')\n        \n        plt.subplot(122)\n        plt.title(\"Generated image\")\n        plt.imshow(generated_sample[0] * 0.5 + 0.5)\n        plt.axis('off')\n        plt.show()\n        \n        \ndef evaluate_cycle(ds, generator_a, generator_b, n_samples=1):\n    fig, axes = plt.subplots(n_samples, 3, figsize=(22, (n_samples*6)))\n    axes = axes.flatten()\n    \n    ds_iter = iter(ds)\n    for n_sample in range(n_samples):\n        idx = n_sample*3\n        example_sample = next(ds_iter)\n        generated_a_sample = generator_a.predict(example_sample)\n        generated_b_sample = generator_b.predict(generated_a_sample)\n        \n        axes[idx].set_title('Input image', fontsize=18)\n        axes[idx].imshow(example_sample[0] * 0.5 + 0.5)\n        axes[idx].axis('off')\n        \n        axes[idx+1].set_title('Generated image', fontsize=18)\n        axes[idx+1].imshow(generated_a_sample[0] * 0.5 + 0.5)\n        axes[idx+1].axis('off')\n        \n        axes[idx+2].set_title('Cycled image', fontsize=18)\n        axes[idx+2].imshow(generated_b_sample[0] * 0.5 + 0.5)\n        axes[idx+2].axis('off')\n        \n    plt.show()\n\ndef create_gif(images_path, gif_path, duration=125):\n    images = []\n    filenames = glob.glob(images_path)\n    filenames.sort(key=lambda x: int(''.join(filter(str.isdigit, x))))\n    for epoch, filename in enumerate(filenames):\n        img = PIL.ImageDraw.Image.open(filename)\n#         fnt = ImageFont.truetype(\"Pillow/Tests/fonts/FreeMono.ttf\", 40)\n        fnt = ImageFont.truetype(\"DejaVuSansMono.ttf\", 15)\n        ImageDraw.Draw(img).text((0, 0),\n                                 f'Epoch {epoch+1}', font=fnt, align='left')\n        images.append(img)\n    imageio.mimsave(gif_path, images, duration=duration) # save gif\n        \n        \n        \n\ndef predict_and_save(input_ds, generator_model, output_path):\n    i = 1\n    for img in input_ds:\n        prediction = generator_model(img, training=False)[0].numpy()\n        prediction = (prediction * 127.5 + 127.5).astype(np.uint8) \n        im = PIL.Image.fromarray(prediction)\n        im.save(f'{output_path}{str(i)}.jpg')\n#         tf.keras.preprocessing.image.save_img(f'{output_path}_photo{str(i)}.jpg', img[0])\n        i += 1\n                \n\ndef downsample(filters, size, apply_instancenorm=True, strides=2):\n    initializer = tf.random_normal_initializer(0., 0.03)\n    gamma_init = tf.keras.initializers.RandomNormal(mean=0.0, stddev=0.03)\n\n    result = tf.keras.Sequential()\n    result.add(L.Conv2D(filters, size, strides=strides, padding='same',\n                        kernel_initializer=initializer, use_bias=False))\n\n    if apply_instancenorm:\n        result.add(tfa.layers.InstanceNormalization(gamma_initializer=gamma_init))\n\n    result.add(L.LeakyReLU())\n\n    return result\n\ndef upsample(filters, size, apply_dropout=False, strides=2):\n    initializer = tf.random_normal_initializer(0., 0.03)\n    gamma_init = tf.keras.initializers.RandomNormal(mean=0.0, stddev=0.03)\n\n    result = tf.keras.Sequential()\n    result.add(L.Conv2DTranspose(filters, size, strides=strides, padding='same',\n                                 kernel_initializer=initializer, use_bias=False))\n\n    result.add(tfa.layers.InstanceNormalization(gamma_initializer=gamma_init))\n\n    if apply_dropout:\n        result.add(L.Dropout(0.6))\n\n    result.add(L.ReLU())\n\n    return result","metadata":{"execution":{"iopub.status.busy":"2023-06-27T00:28:35.971753Z","iopub.execute_input":"2023-06-27T00:28:35.972280Z","iopub.status.idle":"2023-06-27T00:28:36.010082Z","shell.execute_reply.started":"2023-06-27T00:28:35.972248Z","shell.execute_reply":"2023-06-27T00:28:36.009343Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"display_samples(load_dataset(MONET_FILENAMES).batch(1), 4, 6)","metadata":{"execution":{"iopub.status.busy":"2023-06-26T18:10:05.138419Z","iopub.execute_input":"2023-06-26T18:10:05.138682Z","iopub.status.idle":"2023-06-26T18:10:06.824009Z","shell.execute_reply.started":"2023-06-26T18:10:05.138659Z","shell.execute_reply":"2023-06-26T18:10:06.822910Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"display_samples(load_dataset(PHOTO_FILENAMES).batch(1), 4, 6)","metadata":{"execution":{"iopub.status.busy":"2023-06-26T18:10:06.825395Z","iopub.execute_input":"2023-06-26T18:10:06.825727Z","iopub.status.idle":"2023-06-26T18:10:08.335847Z","shell.execute_reply.started":"2023-06-26T18:10:06.825699Z","shell.execute_reply":"2023-06-26T18:10:08.334755Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"\ntransform = models.Sequential([\n  L.RandomRotation((-0.2,0.2),seed=1),\n  L.RandomZoom(height_factor=(-0.15,.15),seed=1),\n  L.RandomTranslation(height_factor=(-0.1, 0.1), width_factor=(-0.15, 0.15), seed=1),\n])","metadata":{"execution":{"iopub.status.busy":"2023-06-26T20:33:32.787236Z","iopub.execute_input":"2023-06-26T20:33:32.787724Z","iopub.status.idle":"2023-06-26T20:33:32.803148Z","shell.execute_reply.started":"2023-06-26T20:33:32.787690Z","shell.execute_reply":"2023-06-26T20:33:32.802103Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"OUTPUT_CHANNELS = 3\n\ndef generator_fn():\n    inputs = L.Input(shape=[HEIGHT, WIDTH, CHANNELS])\n\n    down_stack = [\n        downsample(64, 4, apply_instancenorm=False),\n        downsample(128, 4),                 \n        downsample(256, 4),              \n        downsample(256, 4),                \n        downsample(512, 4),                 \n        downsample(512, 4),                          \n        downsample(512, 4),                     \n        downsample(512, 4),                         \n    ]\n\n    up_stack = [\n        upsample(512, 4, apply_dropout=True), \n        upsample(512, 4, apply_dropout=True),\n        upsample(512, 4, apply_dropout=True), \n        upsample(256, 4),                     \n        upsample(256, 4),                     \n        upsample(128, 4),                   \n        upsample(64, 4),                     \n    ]\n\n    initializer = tf.random_normal_initializer(0., 0.02)\n    last = L.Conv2DTranspose(OUTPUT_CHANNELS, 4,\n                             strides=2,\n                             padding='same',\n                             kernel_initializer=initializer,\n                             activation='relu') \n\n    x = inputs\n    \n    skips = []\n    for down in down_stack:\n        x = down(x)\n        skips.append(x)\n\n    skips = reversed(skips[:-1])\n\n    for up, skip in zip(up_stack, skips):\n        x = up(x)\n        x = L.Concatenate()([x, skip])\n\n    x = last(x)\n\n    return Model(inputs=inputs, outputs=x)","metadata":{"execution":{"iopub.status.busy":"2023-06-27T00:28:44.374743Z","iopub.execute_input":"2023-06-27T00:28:44.375134Z","iopub.status.idle":"2023-06-27T00:28:44.388445Z","shell.execute_reply.started":"2023-06-27T00:28:44.375107Z","shell.execute_reply":"2023-06-27T00:28:44.387453Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def discriminator_fn():\n    initializer = tf.random_normal_initializer(0., 0.02)\n    gamma_init = tf.keras.initializers.RandomNormal(mean=0.0, stddev=0.02)\n\n    inp = L.Input(shape=[HEIGHT, WIDTH, CHANNELS], name='input_image')\n\n    x = inp\n\n    down1 = downsample(64, 4, False)(x) \n    down2 = downsample(128, 4)(down1)\n    down3 = downsample(256, 4)(down2)\n\n    zero_pad1 = L.ZeroPadding2D()(down3)\n    conv = L.Conv2D(512, 4, strides=1,\n                    kernel_initializer=initializer,\n                    use_bias=False)(zero_pad1)\n\n    norm1 = tfa.layers.InstanceNormalization(gamma_initializer=gamma_init)(conv)\n\n    leaky_relu = L.LeakyReLU()(norm1)\n\n    zero_pad2 = L.ZeroPadding2D()(leaky_relu)\n\n    last = L.Conv2D(1, 4, strides=1,\n                    kernel_initializer=initializer)(zero_pad2)\n\n    return Model(inputs=inp, outputs=last)","metadata":{"execution":{"iopub.status.busy":"2023-06-27T00:29:12.632702Z","iopub.execute_input":"2023-06-27T00:29:12.633243Z","iopub.status.idle":"2023-06-27T00:29:12.646086Z","shell.execute_reply.started":"2023-06-27T00:29:12.633196Z","shell.execute_reply":"2023-06-27T00:29:12.644924Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"with strategy.scope():\n    monet_generator = generator_fn()\n    photo_generator = generator_fn() \n\n    monet_discriminator = discriminator_fn() \n    photo_discriminator = discriminator_fn() \n\n\nclass CycleGan(Model):\n    def __init__(\n        self,\n        monet_generator,\n        photo_generator,\n        monet_discriminator,\n        photo_discriminator,\n        lambda_cycle=10,\n    ):\n        super(CycleGan, self).__init__()\n        self.m_gen = monet_generator\n        self.p_gen = photo_generator\n        self.m_disc = monet_discriminator\n        self.p_disc = photo_discriminator\n        self.lambda_cycle = lambda_cycle\n        \n    def compile(\n        self,\n        m_gen_optimizer,\n        p_gen_optimizer,\n        m_disc_optimizer,\n        p_disc_optimizer,\n        gen_loss_fn,\n        disc_loss_fn,\n        cycle_loss_fn,\n        identity_loss_fn\n    ):\n        super(CycleGan, self).compile()\n        self.m_gen_optimizer = m_gen_optimizer\n        self.p_gen_optimizer = p_gen_optimizer\n        self.m_disc_optimizer = m_disc_optimizer\n        self.p_disc_optimizer = p_disc_optimizer\n        self.gen_loss_fn = gen_loss_fn\n        self.disc_loss_fn = disc_loss_fn\n        self.cycle_loss_fn = cycle_loss_fn\n        self.identity_loss_fn = identity_loss_fn\n        \n    def train_step(self, batch_data):\n        real_monet, real_photo = batch_data\n        \n        with tf.GradientTape(persistent=True) as tape:\n            # photo to monet back to photo\n            fake_monet = self.m_gen(real_photo, training=True)\n            cycled_photo = self.p_gen(fake_monet, training=True)\n\n            # monet to photo back to monet\n            fake_photo = self.p_gen(real_monet, training=True)\n            cycled_monet = self.m_gen(fake_photo, training=True)\n\n            # generating itself\n            same_monet = self.m_gen(real_monet, training=True)\n            same_photo = self.p_gen(real_photo, training=True)\n\n            # discriminator used to check, inputing real images\n            disc_real_monet = self.m_disc(real_monet, training=True)\n            disc_real_photo = self.p_disc(real_photo, training=True)\n\n            # discriminator used to check, inputing fake images\n            disc_fake_monet = self.m_disc(fake_monet, training=True)\n            disc_fake_photo = self.p_disc(fake_photo, training=True)\n\n            # evaluates generator loss\n            monet_gen_loss = self.gen_loss_fn(disc_fake_monet)\n            photo_gen_loss = self.gen_loss_fn(disc_fake_photo)\n\n            # evaluates total cycle consistency loss\n            total_cycle_loss = self.cycle_loss_fn(real_monet, cycled_monet, self.lambda_cycle) + self.cycle_loss_fn(real_photo, cycled_photo, self.lambda_cycle)\n\n            # evaluates total generator loss\n            total_monet_gen_loss = monet_gen_loss + total_cycle_loss + self.identity_loss_fn(real_monet, same_monet, self.lambda_cycle)\n            total_photo_gen_loss = photo_gen_loss + total_cycle_loss + self.identity_loss_fn(real_photo, same_photo, self.lambda_cycle)\n\n            # evaluates discriminator loss\n            monet_disc_loss = self.disc_loss_fn(disc_real_monet, disc_fake_monet)\n            photo_disc_loss = self.disc_loss_fn(disc_real_photo, disc_fake_photo)\n\n        # Calculate the gradients for generator and discriminator\n        monet_generator_gradients = tape.gradient(total_monet_gen_loss,\n                                                  self.m_gen.trainable_variables)\n        photo_generator_gradients = tape.gradient(total_photo_gen_loss,\n                                                  self.p_gen.trainable_variables)\n\n        monet_discriminator_gradients = tape.gradient(monet_disc_loss,\n                                                      self.m_disc.trainable_variables)\n        photo_discriminator_gradients = tape.gradient(photo_disc_loss,\n                                                      self.p_disc.trainable_variables)\n\n        self.m_gen_optimizer.apply_gradients(zip(monet_generator_gradients,\n                                                 self.m_gen.trainable_variables))\n\n        self.p_gen_optimizer.apply_gradients(zip(photo_generator_gradients,\n                                                 self.p_gen.trainable_variables))\n\n        self.m_disc_optimizer.apply_gradients(zip(monet_discriminator_gradients,\n                                                  self.m_disc.trainable_variables))\n\n        self.p_disc_optimizer.apply_gradients(zip(photo_discriminator_gradients,\n                                                  self.p_disc.trainable_variables))\n        \n        return {\n            'monet_gen_loss': total_monet_gen_loss,\n            'photo_gen_loss': total_photo_gen_loss,\n            'monet_disc_loss': monet_disc_loss,\n            'photo_disc_loss': photo_disc_loss\n        }","metadata":{"execution":{"iopub.status.busy":"2023-06-27T00:29:24.308610Z","iopub.execute_input":"2023-06-27T00:29:24.308967Z","iopub.status.idle":"2023-06-27T00:29:30.627327Z","shell.execute_reply.started":"2023-06-27T00:29:24.308940Z","shell.execute_reply":"2023-06-27T00:29:30.626247Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"%pip install ann_visualizer\nfrom ann_visualizer.visualize import ann_viz;","metadata":{"execution":{"iopub.status.busy":"2023-06-27T00:29:35.644730Z","iopub.execute_input":"2023-06-27T00:29:35.645093Z","iopub.status.idle":"2023-06-27T00:29:41.701476Z","shell.execute_reply.started":"2023-06-27T00:29:35.645056Z","shell.execute_reply":"2023-06-27T00:29:41.700247Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# keras visualizer here (ugly output ;/)\n# my_settings = []\n# visualizer(monet_discriminator, file_name='monet_disc', file_format='png', view=False, settings=None)\n# ann_viz(monet_generator, title=\"Monet_generator\")","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"with strategy.scope():\n    # Discriminator loss {0: fake, 1: real} (The discriminator loss outputs the average of the real and generated loss)\n    def discriminator_loss(real, generated):\n        real_loss = losses.BinaryCrossentropy(from_logits=True, reduction=losses.Reduction.NONE)(tf.ones_like(real), real)\n\n        generated_loss = losses.BinaryCrossentropy(from_logits=True, reduction=losses.Reduction.NONE)(tf.zeros_like(generated), generated)\n\n        total_disc_loss = real_loss + generated_loss\n\n        return total_disc_loss * 0.5\n    \n    # Generator loss\n    def generator_loss(generated):\n        return losses.BinaryCrossentropy(from_logits=True, reduction=losses.Reduction.NONE)(tf.ones_like(generated), generated)\n    \n    \n    # Cycle consistency loss (measures if original photo and the twice transformed photo to be similar to one another)\n    with strategy.scope():\n        def calc_cycle_loss(real_image, cycled_image, LAMBDA):\n            loss1 = tf.reduce_mean(tf.abs(real_image - cycled_image))\n\n            return LAMBDA * loss1\n\n    # Identity loss (compares the image with its generator (i.e. photo with photo generator))\n    with strategy.scope():\n        def identity_loss(real_image, same_image, LAMBDA):\n            loss = tf.reduce_mean(tf.abs(real_image - same_image))\n            return LAMBDA * 0.5 * loss","metadata":{"execution":{"iopub.status.busy":"2023-06-26T18:19:03.700314Z","iopub.execute_input":"2023-06-26T18:19:03.700722Z","iopub.status.idle":"2023-06-26T18:19:03.710862Z","shell.execute_reply.started":"2023-06-26T18:19:03.700692Z","shell.execute_reply":"2023-06-26T18:19:03.709828Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"monet_ds = get_dataset(MONET_FILENAMES, augment=transform, batch_size=BATCH_SIZE)\nphoto_ds = get_dataset(PHOTO_FILENAMES, augment=transform, batch_size=BATCH_SIZE)\ngan_ds = tf.data.Dataset.zip((monet_ds, photo_ds))\n\nphoto_ds_eval = get_dataset(PHOTO_FILENAMES, repeat=False, shuffle=False, batch_size=1)\nmonet_ds_eval = get_dataset(MONET_FILENAMES, repeat=False, shuffle=False, batch_size=1)\n\nclass GANCallback(Callback):\n    def __init__(self, num_img=1, monet_path='monet', photo_path='photo'):\n        self.num_img = num_img\n        self.monet_path = monet_path\n        self.photo_path = photo_path\n        # Create directories to save the generate images\n        if not os.path.exists(self.monet_path):\n            os.makedirs(self.monet_path)\n        if not os.path.exists(self.photo_path):\n            os.makedirs(self.photo_path)\n\n    def on_epoch_end(self, epoch, logs=None):\n        for i, img in enumerate(photo_ds_eval.take(self.num_img)):\n            prediction = monet_generator(img, training=False)[0].numpy()\n            prediction = (prediction * 127.5 + 127.5).astype(np.uint8)\n            prediction = PIL.Image.fromarray(prediction)\n            prediction.save(f'{self.monet_path}/generated_{i}_{epoch+1}.png')\n            \n        for i, img in enumerate(monet_ds_eval.take(self.num_img)):\n            prediction = photo_generator(img, training=False)[0].numpy()\n            prediction = (prediction * 127.5 + 127.5).astype(np.uint8)\n            prediction = PIL.Image.fromarray(prediction)\n            prediction.save(f'{self.photo_path}/generated_{i}_{epoch+1}.png')\n","metadata":{"execution":{"iopub.status.busy":"2023-06-26T20:32:29.499991Z","iopub.execute_input":"2023-06-26T20:32:29.500447Z","iopub.status.idle":"2023-06-26T20:32:30.674539Z","shell.execute_reply.started":"2023-06-26T20:32:29.500416Z","shell.execute_reply":"2023-06-26T20:32:30.673239Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"display_samples(monet_ds, 2, 3)","metadata":{"execution":{"iopub.status.busy":"2023-06-26T20:32:36.918904Z","iopub.execute_input":"2023-06-26T20:32:36.919946Z","iopub.status.idle":"2023-06-26T20:32:38.648644Z","shell.execute_reply.started":"2023-06-26T20:32:36.919904Z","shell.execute_reply":"2023-06-26T20:32:38.647136Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"display_samples(photo_ds, 2, 3)","metadata":{"execution":{"iopub.status.busy":"2023-06-26T20:32:50.038071Z","iopub.execute_input":"2023-06-26T20:32:50.038520Z","iopub.status.idle":"2023-06-26T20:32:51.603351Z","shell.execute_reply.started":"2023-06-26T20:32:50.038489Z","shell.execute_reply":"2023-06-26T20:32:51.601893Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"with strategy.scope():\n    # Create generators\n    monet_generator_optimizer = optimizers.Adam(8e-5, beta_1=0.6)\n    photo_generator_optimizer = optimizers.Adam(8e-5, beta_1=0.6)\n\n    # Create discriminators\n    monet_discriminator_optimizer = optimizers.Adam(3e-4, beta_1=0.6)\n    photo_discriminator_optimizer = optimizers.Adam(3e-4, beta_1=0.6)\n    \n#     monet_generator.load_weights('/kaggle/working/checkpoints/monet_generator_epochs7')\n#     photo_generator.load_weights('/kaggle/working/checkpoints/photo_generator_epochs7')\n#     monet_discriminator.load_weights('/kaggle/working/checkpoints/monet_discriminator_epochs7')\n#     photo_discriminator.load_weights('/kaggle/working/checkpoints/photo_discriminator_epochs7')\n#     monet_generator.load('/kaggle/working/model/monet_generator_epochs7.h5')\n#     photo_generator.load('/kaggle/working/model/photo_generator_epochs7.h5')\n#     monet_discriminator.load('/kaggle/working/model/monet_discriminator_epochs7.h5')\n#     photo_discriminator.load('/kaggle/working/model/photo_discriminator_epochs7.h5')\n    \n    # Create GAN\n    gan_model = CycleGan(monet_generator, photo_generator, \n                         monet_discriminator, photo_discriminator)\n\n    gan_model.compile(m_gen_optimizer=monet_generator_optimizer,\n                      p_gen_optimizer=photo_generator_optimizer,\n                      m_disc_optimizer=monet_discriminator_optimizer,\n                      p_disc_optimizer=photo_discriminator_optimizer,\n                      gen_loss_fn=generator_loss,\n                      disc_loss_fn=discriminator_loss,\n                      cycle_loss_fn=calc_cycle_loss,\n                      identity_loss_fn=identity_loss)\n    \n\nhistory = gan_model.fit(gan_ds,\n                        epochs=120,\n                        callbacks=[GANCallback()], \n                        steps_per_epoch=(n_monet_samples//BATCH_SIZE), \n                        verbose=2).history","metadata":{"execution":{"iopub.status.busy":"2023-06-26T20:33:44.768243Z","iopub.execute_input":"2023-06-26T20:33:44.769293Z","iopub.status.idle":"2023-06-26T22:03:43.020067Z","shell.execute_reply.started":"2023-06-26T20:33:44.769251Z","shell.execute_reply":"2023-06-26T22:03:43.018590Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# monet_generator.save_weights('/kaggle/working/checkpoints/monet_generator_epochs7')\n# photo_generator.save_weights('/kaggle/working/checkpoints/photo_generator_epochs7')\n# monet_discriminator.save_weights('/kaggle/working/checkpoints/monet_discriminator_epochs7')\n# photo_discriminator.save_weights('/kaggle/working/checkpoints/photo_discriminator_epochs7')\n# monet_generator.save('/kaggle/working/model/monet_generator_epochs7.h5')\n# photo_generator.save('/kaggle/working/model/photo_generator_epochs7.h5')\n# monet_discriminator.save('/kaggle/working/model/monet_discriminator_epochs7.h5')\n# photo_discriminator.save('/kaggle/working/model/photo_discriminator_epochs7.h5')","metadata":{"execution":{"iopub.status.busy":"2023-06-26T18:10:25.126445Z","iopub.status.idle":"2023-06-26T18:10:25.126766Z","shell.execute_reply.started":"2023-06-26T18:10:25.126602Z","shell.execute_reply":"2023-06-26T18:10:25.126617Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# for k in history.history:\n#   plt.plot(history.history[k],label=k)\n# plt.legend()\n# dir(history)","metadata":{"execution":{"iopub.status.busy":"2023-06-26T20:11:43.533690Z","iopub.execute_input":"2023-06-26T20:11:43.534689Z","iopub.status.idle":"2023-06-26T20:11:43.539515Z","shell.execute_reply.started":"2023-06-26T20:11:43.534635Z","shell.execute_reply":"2023-06-26T20:11:43.538156Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"display_generated_samples(photo_ds_eval.take(20), monet_generator, 20)","metadata":{"execution":{"iopub.status.busy":"2023-06-26T23:09:31.917275Z","iopub.execute_input":"2023-06-26T23:09:31.917683Z","iopub.status.idle":"2023-06-26T23:09:48.568373Z","shell.execute_reply.started":"2023-06-26T23:09:31.917649Z","shell.execute_reply":"2023-06-26T23:09:48.566893Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"create_gif('/kaggle/working/monet/generated_13*.png', 'monet2.gif', duration=250)","metadata":{"execution":{"iopub.status.busy":"2023-06-26T23:52:57.578311Z","iopub.execute_input":"2023-06-26T23:52:57.578804Z","iopub.status.idle":"2023-06-26T23:53:28.235923Z","shell.execute_reply.started":"2023-06-26T23:52:57.578773Z","shell.execute_reply":"2023-06-26T23:53:28.234623Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"create_gif('/kaggle/working/photo/generated_21*.png', 'photo2.gif', duration=250)","metadata":{"execution":{"iopub.status.busy":"2023-06-26T23:53:28.237837Z","iopub.execute_input":"2023-06-26T23:53:28.238149Z","iopub.status.idle":"2023-06-26T23:53:54.789146Z","shell.execute_reply.started":"2023-06-26T23:53:28.238122Z","shell.execute_reply":"2023-06-26T23:53:54.787949Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"evaluate_cycle(photo_ds_eval.take(5), monet_generator, photo_generator, n_samples=5)","metadata":{"execution":{"iopub.status.busy":"2023-06-26T22:10:23.189365Z","iopub.execute_input":"2023-06-26T22:10:23.190148Z","iopub.status.idle":"2023-06-26T22:10:33.033595Z","shell.execute_reply.started":"2023-06-26T22:10:23.190112Z","shell.execute_reply":"2023-06-26T22:10:33.032271Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"evaluate_cycle(monet_ds_eval.take(5), photo_generator, monet_generator, n_samples=5)","metadata":{"execution":{"iopub.status.busy":"2023-06-26T22:11:11.313819Z","iopub.execute_input":"2023-06-26T22:11:11.314273Z","iopub.status.idle":"2023-06-26T22:11:21.357053Z","shell.execute_reply.started":"2023-06-26T22:11:11.314242Z","shell.execute_reply":"2023-06-26T22:11:21.355778Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"%%time\nos.makedirs('../images2/') # Create folder to save generated images\n\npredict_and_save(photo_ds_eval, monet_generator, '../images2/')","metadata":{"execution":{"iopub.status.busy":"2023-06-26T22:42:48.951539Z","iopub.execute_input":"2023-06-26T22:42:48.952026Z","iopub.status.idle":"2023-06-26T23:09:31.915129Z","shell.execute_reply.started":"2023-06-26T22:42:48.951991Z","shell.execute_reply":"2023-06-26T23:09:31.913731Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"shutil.make_archive('/kaggle/working/images2', 'zip', '../images2/')\n\nprint(f\"Generated samples: {len([name for name in os.listdir('../images2/') if os.path.isfile(os.path.join('../images2/', name))])}\")","metadata":{"execution":{"iopub.status.busy":"2023-06-26T23:09:48.570085Z","iopub.execute_input":"2023-06-26T23:09:48.570494Z","iopub.status.idle":"2023-06-26T23:09:52.577516Z","shell.execute_reply.started":"2023-06-26T23:09:48.570461Z","shell.execute_reply":"2023-06-26T23:09:52.576093Z"},"trusted":true},"execution_count":null,"outputs":[]}]}