2_Regression_Baseline_Model_Comparison.ipynb

{"cells":[{"cell_type":"markdown","metadata":{"id":"GmBY0QElpB9n"},"source":["\n","# Prepare Moduls, Function, Path, Data\n"]},{"cell_type":"markdown","metadata":{"id":"neuR_JGoSki5"},"source":["## Imports"]},{"cell_type":"markdown","metadata":{"id":"CN790fN9pAMp"},"source":[]},{"cell_type":"code","execution_count":1,"metadata":{"id":"yVusG7IwSfWU","executionInfo":{"status":"ok","timestamp":1707289403900,"user_tz":-60,"elapsed":4196,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["from tensorflow.keras.preprocessing.image import ImageDataGenerator\n","from google.colab import drive\n","import os\n","import pandas as pd\n","import numpy as np\n","import matplotlib.pyplot as plt\n","from sklearn.model_selection import train_test_split\n","from abc import ABC, abstractmethod\n","import pandas as pd\n","\n","from keras import backend as K\n","from tensorflow.python.keras.utils.layer_utils import count_params\n","\n","from tensorflow.keras.applications import resnet50\n","import tensorflow as tf\n","from tensorflow.keras.layers import (\n","    Dense,\n","    Dropout,\n","    Flatten,\n",")\n","from tensorflow.keras.models import Model"]},{"cell_type":"markdown","metadata":{"id":"7qM_UgFLSqcF"},"source":["##PEP8 standard"]},{"cell_type":"code","execution_count":3,"metadata":{"id":"n3jjfuPNSrY4","executionInfo":{"status":"ok","timestamp":1707289424474,"user_tz":-60,"elapsed":20579,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["%%capture\n","try:\n","    from nbpep8.nbpep8 import pep8\n","except ModuleNotFoundError:\n","    !pip install pycodestyle\n","    !pip install --index-url https://test.pypi.org/simple/ nbpep8\n","from nbpep8.nbpep8 import pep8"]},{"cell_type":"markdown","metadata":{"id":"Dr5iOQgfStF9"},"source":["## Custom packages"]},{"cell_type":"code","execution_count":4,"metadata":{"id":"cv1XS4e8SsSj","executionInfo":{"status":"ok","timestamp":1707289424904,"user_tz":-60,"elapsed":433,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["%%capture\n","def download_files(url_dict):\n","    for file, url in url_dict.items():\n","        print(f'Downloading {file}')\n","        !wget -O {file} {url} {file}\n","\n","git_download_path = \\\n","'https://raw.githubusercontent.com/PashaIanko/Sklearn-Utils/main/'\n","\n","files = [\n","    'path_manager.py',\n","]\n","\n","url_dict = {file: git_download_path + file for file in files}\n","download_files(url_dict)"]},{"cell_type":"code","execution_count":5,"metadata":{"id":"_RDScb8wSvVT","executionInfo":{"status":"ok","timestamp":1707289424904,"user_tz":-60,"elapsed":3,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["import importlib\n","import path_manager\n","\n","def reload_all(modules_list_):\n","    for module in modules_list_:\n","        importlib.reload(module)\n","\n","reload_all(\n","    [\n","        path_manager,\n","\n","    ]\n",")\n","\n","from path_manager import PathManager"]},{"cell_type":"markdown","metadata":{"id":"ZkKxmfS9SznR"},"source":["## Setup paths"]},{"cell_type":"code","execution_count":6,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":46601,"status":"ok","timestamp":1707289471503,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"},"user_tz":-60},"id":"CGZstgihSwbl","outputId":"1aae6996-8820-4ce5-e532-375876f912f6"},"outputs":[{"output_type":"stream","name":"stdout","text":["Mounted at /content/gdrive\n","/content/gdrive/MyDrive/CrowdCounting/Data/preproc_trial_1 already exists\n","/content/gdrive/MyDrive/CrowdCounting/Models/trial_1 already exists\n"]}],"source":["drive.mount('/content/gdrive', force_remount=True)\n","gdrive_path = '/content/gdrive/MyDrive/'\n","\n","PREPROC_TRIAL = 1\n","MODELS_TRIAL = 1\n","COMPETITION_PATH = gdrive_path + 'CrowdCounting/'\n","\n","manager = PathManager(\n","    competition_path=COMPETITION_PATH,\n","    preprocessing_trial=PREPROC_TRIAL,\n","    models_trial=MODELS_TRIAL\n",")\n","\n","manager.setup_paths()"]},{"cell_type":"markdown","metadata":{"id":"EEhgFmlqS9at"},"source":["## Helper functions"]},{"cell_type":"code","execution_count":7,"metadata":{"id":"z-pCnSLMS6hH","executionInfo":{"status":"ok","timestamp":1707289471504,"user_tz":-60,"elapsed":10,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["def count_parameters(model):\n","    trainable_count = count_params(model.trainable_weights)\n","    non_trainable_count = count_params(model.non_trainable_weights)\n","\n","    return {\n","        'Total parameters': trainable_count + non_trainable_count,\n","        'Trainable parameters': trainable_count,\n","        'Non-trainable parameters': non_trainable_count\n","    }"]},{"cell_type":"code","execution_count":8,"metadata":{"id":"dcfl1oP6noEB","executionInfo":{"status":"ok","timestamp":1707289471945,"user_tz":-60,"elapsed":448,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["def collect_comparison_data(models_dict_):\n","    validation_results = {}\n","    model_parameters = {}\n","\n","    for model_name, model_object in MODELS.items():\n","        print(f'Working with {model_name}')\n","\n","        model_object.prepare_model()\n","        model_object.compile_model()\n","\n","        model_object.model.load_weights(\n","            os.path.join(\n","                manager.models_trial_path,\n","                model_object.name,\n","                model_object.name\n","            )\n","        )\n","        validation_results[model_name] = model_object.model.evaluate(\n","            model_object.validation_generator\n","        )\n","\n","        model_parameters[model_name] = count_parameters(model_object.model)\n","\n","    return {\n","        'validation_results': validation_results,\n","        'model_parameters': model_parameters\n","    }"]},{"cell_type":"code","execution_count":9,"metadata":{"id":"yNupQ1isjNC1","executionInfo":{"status":"ok","timestamp":1707289471945,"user_tz":-60,"elapsed":9,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["def plot_validation_results(validation_performance_dict_):\n","    mae_dict = {\n","        model_name: [data[2]] \\\n","        for model_name, data in validation_performance_dict_.items()\n","    }\n","\n","    df = pd.DataFrame(mae_dict, index=None)\n","    df = df.T\n","    print(df)\n","    df = df.sort_values(by=0)\n","\n","    fontsize = 14\n","    labelsize = 14\n","    ax = df.plot(kind='bar')\n","    ax.set_ylabel('Validation MAE', fontsize=fontsize)\n","    ax.set_xlabel('Model', fontsize=fontsize)\n","    ax.set_title('Validation performance of baseline models', fontsize=fontsize)\n","    ax.legend(prop={'size': fontsize})\n","    ax.tick_params(labelsize=labelsize)\n","    ax.get_legend().remove()\n","    ax.grid()\n","    plt.xticks(rotation=45)"]},{"cell_type":"code","execution_count":10,"metadata":{"id":"QgXJJmmIjSpN","executionInfo":{"status":"ok","timestamp":1707289471945,"user_tz":-60,"elapsed":8,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["def plot_model_parameters(parameters_dict_):\n","    total_parameters_dict = {\n","        model_name: [model_dict['Total parameters']] \\\n","        for model_name, model_dict in parameters_dict_.items()\n","    }\n","\n","    trainable_parameters_dict = {\n","        model_name: [model_dict['Trainable parameters']] \\\n","        for model_name, model_dict in parameters_dict_.items()\n","    }\n","\n","    _, ax = plt.subplots(1, 2, figsize=(14, 8))\n","\n","    fontsize = 14\n","    labelsize = 14\n","\n","    df_total_parameters = pd.DataFrame(total_parameters_dict)\n","    df_total_parameters = df_total_parameters.T\n","    df_total_parameters.sort_values(by=0, inplace=True)\n","    df_total_parameters.plot(kind='bar', ax=ax[0], fontsize=fontsize)\n","    ax[0].set_title('Total number of model parameters', fontsize=fontsize)\n","    ax[0].grid()\n","    ax[0].get_legend().remove()\n","\n","    df_trainable_parameters = pd.DataFrame(trainable_parameters_dict)\n","    df_trainable_parameters = df_trainable_parameters.T\n","    df_trainable_parameters.sort_values(by=0, inplace=True)\n","    df_trainable_parameters.plot(kind='bar', ax=ax[1], fontsize = 14)\n","\n","\n","\n","    ax[1].set_title(\n","        'Number of trainable parameters' \\\n","        + '\\nfor baseline ' \\\n","        + 'models\\n(with 7 top layers unfreezed)',\n","        fontsize=fontsize\n","    )\n","    ax[1].grid()\n","    ax[1].get_legend().remove()\n","    plt.tick_params(rotation=45, axis='x')\n","\n","    plt.tight_layout()"]},{"cell_type":"code","execution_count":11,"metadata":{"id":"gIIEGJAFjaWR","executionInfo":{"status":"ok","timestamp":1707289471945,"user_tz":-60,"elapsed":8,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["def compare_learning_curve(\n","    csv_dict_,\n","    metrics_\n","):\n","\n","    fontsize = 20\n","    legendsize = 16\n","\n","    _, ax = plt.subplots(figsize=(10, 7))\n","    for model_name, csv_filepath in csv_dict_.items():\n","        assert os.path.isfile(csv_filepath)\n","        df = pd.read_csv(csv_filepath)\n","        assert metrics_ in df.columns\n","\n","        data = df[metrics_]\n","        ax.plot(data, label=model_name, marker='o', markersize=4)\n","    ax.legend(prop={'size': legendsize})\n","    ax.grid()\n","\n","    ax.set_xlabel('Epoch', fontsize=fontsize)\n","    ax.set_ylabel(metrics_, fontsize=fontsize)\n","    ax.set_title(f'{metrics_} vs Epoch for Baseline Models', fontsize=fontsize)\n","\n","    ax.tick_params(labelsize=17)"]},{"cell_type":"code","execution_count":12,"metadata":{"id":"cpYBaFvkjilw","executionInfo":{"status":"ok","timestamp":1707289471946,"user_tz":-60,"elapsed":8,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["def visualize_training_times(training_times_dict_):\n","    _, ax = plt.subplots(1, 2, figsize=(16, 4))\n","    fontsize=16\n","    legendsize=11\n","    labelsize=14\n","\n","    for model_name, training_time in training_times_dict_.items():\n","        ax[0].plot(\n","            training_time[1:],\n","            label=model_name,\n","            marker='o',\n","            markersize=5\n","            # linestyle='--'\n","        )\n","    ax[0].grid()\n","    ax[0].set_title(\n","        'Time per epoch for baseline models' \\\n","        + '\\n(Starting from 2nd epoch)',\n","        fontsize=fontsize\n","    )\n","    ax[0].set_ylabel('Time, sec', fontsize=fontsize)\n","    ax[0].set_xlabel(f'Epoch', fontsize=fontsize)\n","    ax[0].legend(prop={'size': legendsize})\n","    ax[0].tick_params(labelsize=labelsize)\n","    # bbox_to_anchor=(1.1, 1.05)\n","\n","\n","    first_epoch_dt_dict = {\n","        model_name: [times[0]] \\\n","        for model_name, times in training_times_dict_.items()\n","    }\n","    df = pd.DataFrame(first_epoch_dt_dict)\n","    df = df.T\n","    df.sort_values(by=0, inplace=True)\n","    df.plot(kind='bar', ax=ax[1])\n","    ax[1].get_legend().remove()\n","    ax[1].tick_params(labelrotation=45, axis='x')\n","    ax[1].grid()\n","    ax[1].set_title(f'Time of the first epoch', fontsize=fontsize)\n","    ax[1].set_ylabel(f'Time, sec', fontsize=fontsize)\n","    ax[1].set_xlabel(f'Model', fontsize=fontsize)\n","    ax[1].tick_params(labelsize=labelsize)"]},{"cell_type":"code","execution_count":13,"metadata":{"id":"Rrcp8hiQjrYh","executionInfo":{"status":"ok","timestamp":1707289471946,"user_tz":-60,"elapsed":8,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["def plot_validation_performance_vs_unfreezed_layers(metrics_):\n","    validation_performances = {}\n","    for model_name, csv_filepath in csv_files_dict.items():\n","        n_unfreezed_layers = int(model_name.split('_')[1])\n","        learning_curve_df = pd.read_csv(\n","            os.path.join(manager.models_trial_path, csv_filepath))\n","        final_metrics = learning_curve_df[metrics_].values[-1]\n","        validation_performances[n_unfreezed_layers] = [final_metrics]\n","\n","    df = pd.DataFrame(validation_performances)\n","    df = df.T\n","    fontsize = 14\n","    labelsize = 12\n","    # df.sort_values(by=0, inplace=True)\n","    ax = df.plot(kind='bar')\n","    ax.set_xlabel(f'Number of unfrozen layers', fontsize=fontsize)\n","    ax.set_ylabel(f'Validation performance (MAE)', fontsize=fontsize)\n","    ax.tick_params(labelsize=labelsize, labelrotation=0)\n","\n","    ax.get_legend().remove()\n","    ax.grid()\n","    ax.set_title(f'Validation performance VS number\\nof unfreezed layers for Xception architecture', fontsize=fontsize)"]},{"cell_type":"code","execution_count":14,"metadata":{"id":"ao6xjj2Qjzpu","executionInfo":{"status":"ok","timestamp":1707289471946,"user_tz":-60,"elapsed":7,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["def plot_number_of_trainable_parameters(\n","    model_class_,\n","    model_params_dict_,\n","    n_unfreezed_layers_list_\n","):\n","    fontsize = 14\n","    labelsize = 12\n","    plot_data = {}\n","    for n_unfreezed_layers in n_unfreezed_layers_list_:\n","        model = model_class_(\n","            **model_params_dict_\n","        )\n","        model.prepare_model()\n","        model.unfreeze_n_layers(n_unfreezed_layers)\n","        params_dict = count_parameters(model.model)\n","        trainable_parameters = params_dict['Trainable parameters']\n","        plot_data[n_unfreezed_layers] = [trainable_parameters]\n","\n","    df = pd.DataFrame(plot_data).T\n","    ax = df.plot(kind='bar', fontsize=fontsize)\n","    ax.get_legend().remove()\n","    ax.grid()\n","    ax.tick_params(labelrotation=0)\n","    ax.set_title(f'Number of trainable parameters \\nVS unfreezed layers', fontsize=fontsize)\n","    ax.set_xlabel(f'Number of unfreezed layers (from top)', fontsize=fontsize)\n","    ax.set_ylabel(f'Number of trainable parameters', fontsize=fontsize)"]},{"cell_type":"code","execution_count":15,"metadata":{"id":"kQ1nxibjj7oE","executionInfo":{"status":"ok","timestamp":1707289471946,"user_tz":-60,"elapsed":7,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["def plot_history(csv_filepaths_dict_, metrics_, title_=''):\n","    _, ax = plt.subplots()\n","\n","    fontsize = 14\n","    legendsize = 11\n","    labelsize = 12\n","\n","    for model_name, path in csv_filepaths_dict_.items():\n","        assert os.path.isfile(path)\n","        df = pd.read_csv(path)\n","        ax.plot(df[metrics_], label=f'{model_name} {metrics_}', marker='o', markersize=2)\n","        ax.plot(df[f'val_{metrics_}'], label=f'{model_name} val_{metrics_}', marker='o', markersize=2)\n","\n","    ax.set_ylabel(f'Learning curve for {metrics_}', fontsize=fontsize)\n","    ax.set_xlabel(f'Epoch', fontsize=fontsize)\n","    ax.set_title(title_)\n","    ax.tick_params(labelsize=labelsize)\n","    ax.grid()\n","    ax.legend(prop={'size':legendsize})"]},{"cell_type":"markdown","metadata":{"id":"fWzz3eYVTC9j"},"source":["## Data split"]},{"cell_type":"code","execution_count":16,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":517,"status":"ok","timestamp":1707289472457,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"},"user_tz":-60},"id":"-5KTBAI5TD0q","outputId":"fc3fc08f-d690-48a4-e30f-cf8d048ed305"},"outputs":[{"output_type":"stream","name":"stdout","text":["   id  count      image_name\n","0   1     35  seq_000001.jpg\n"]}],"source":["frames_directory = os.path.join(manager.data_root_path, 'frames')\n","labels_df = pd.read_csv(os.path.join(manager.data_root_path, 'labels.csv'))\n","labels_df['image_name'] = labels_df['id'].map('seq_{:06d}.jpg'.format)\n","print(labels_df.head(1))"]},{"cell_type":"code","execution_count":17,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":9,"status":"ok","timestamp":1707289472457,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"},"user_tz":-60},"id":"6eQnulCETFGW","outputId":"200affac-265e-4368-cfcc-5c27edc6be11"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["((1600, 3), (400, 3))"]},"metadata":{},"execution_count":17}],"source":["TRAINVAL_SIZE = 0.8  # 0.8\n","RANDOM_STATE = 42\n","\n","trainval_labels_df, test_labels_df = train_test_split(\n","    labels_df,\n","    train_size=TRAINVAL_SIZE,\n","    random_state=RANDOM_STATE,\n",")\n","trainval_labels_df.shape, test_labels_df.shape"]},{"cell_type":"markdown","metadata":{"id":"juSMgE8UTZ6B"},"source":["# Classes definition"]},{"cell_type":"code","execution_count":18,"metadata":{"id":"vjE28uf3TQHa","executionInfo":{"status":"ok","timestamp":1707289472458,"user_tz":-60,"elapsed":7,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["class NNModel(ABC):\n","    def __init__(\n","        self,\n","        name_,\n","        input_image_size_,\n","        validation_split_,\n","        batch_size_,\n","        trainval_labels_df_,\n","        test_labels_df_\n","    ):\n","        self.random_state = 42\n","        self.name = name_\n","        self.input_image_size = input_image_size_\n","        self.validation_split = validation_split_\n","        self.batch_size = batch_size_\n","\n","\n","        self.data_generator_parameters = None\n","        self.train_generator = None\n","        self.validation_generator = None\n","        self.test_generator = None\n","        self.model = None\n","        self.init_data_generator_parameters()\n","        self.init_data_generators(\n","            trainval_labels_df_,\n","            test_labels_df_\n","        )\n","\n","    def __repr__(self):\n","        return f'''\n","            {\n","                self.random_state,\n","                self.name,\n","                self.input_image_size,\n","                self.validation_split,\n","                self.batch_size\n","            }\n","        '''\n","    @abstractmethod\n","    def init_data_generator_parameters(self):\n","        # Augmentation / preprocessing parameters\n","        pass\n","\n","    def init_data_generators(self, trainval_labels_df_, test_labels_df_):\n","        trainval_data_generator = ImageDataGenerator(\n","            **self.data_generator_parameters,\n","            validation_split=self.validation_split)\n","\n","        test_data_generator = ImageDataGenerator(\n","            **self.data_generator_parameters)\n","\n","        common_flow_params = dict(\n","            directory=frames_directory,\n","            x_col='image_name',\n","            y_col='count',\n","            target_size=self.input_image_size,\n","            color_mode='rgb',\n","            class_mode='raw',\n","            batch_size=self.batch_size,\n","            shuffle=True,\n","            seed=self.random_state\n","        )\n","\n","        self.train_generator = trainval_data_generator.flow_from_dataframe(\n","            **common_flow_params,\n","            subset='training',\n","            dataframe=trainval_labels_df_\n","        )\n","\n","        self.validation_generator = trainval_data_generator.flow_from_dataframe(\n","            **common_flow_params,\n","            subset='validation',\n","            dataframe=trainval_labels_df_\n","        )\n","\n","        self.test_generator = test_data_generator.flow_from_dataframe(\n","            **common_flow_params,\n","            dataframe=test_labels_df_\n","        )\n","\n","    def visualize_data(self):\n","        X_batch, Y_batch = next(self.train_generator)\n","        print(X_batch.shape, np.max(X_batch), np.min(X_batch))\n","\n","        _, ax = plt.subplots(3, 3)\n","        ax = ax.flatten()\n","        for i in range(9):\n","            ax[i].imshow(X_batch[i])\n","            ax[i].set_title(Y_batch[i])\n","            ax[i].axis('off')\n","\n","        _, ax_hist = plt.subplots()\n","        ax.hist.hist(Y_batch)\n","\n","    @abstractmethod\n","    def prepare_model(self):\n","        pass\n","\n","    def compile_model(self):\n","        self.model.compile(\n","            optimizer=tf.keras.optimizers.legacy.Adam(learning_rate=0.001),\n","            loss='mse',\n","            metrics=['mse', 'mae']\n","        )\n","\n","    def unfreeze_n_layers(self, n_layers_to_freeze_):\n","        assert n_layers_to_freeze_ > 0\n","\n","        for layer in self.model.layers[:-n_layers_to_freeze_]:\n","            layer.trainable = False\n","        for layer in self.model.layers[-n_layers_to_freeze_:]:\n","            layer.trainable = True\n","\n","    def fit(self, epochs_, callbacks_):\n","        self.learning_history = self.model.fit_generator(\n","            self.train_generator,\n","            validation_data=self.validation_generator,\n","            epochs=epochs_,\n","            callbacks=callbacks_\n","        )\n","        return self.learning_history"]},{"cell_type":"markdown","metadata":{"id":"9jWjvkNfo_Qy"},"source":[]},{"cell_type":"markdown","metadata":{"id":"tMpMtJJpdGRR"},"source":["## Resnet"]},{"cell_type":"code","execution_count":19,"metadata":{"id":"QDhcM5UCTkkE","executionInfo":{"status":"ok","timestamp":1707289472458,"user_tz":-60,"elapsed":7,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["class ResNetModel(NNModel):\n","    def __init__(self, *args, **kwargs):\n","        super().__init__(*args, **kwargs)\n","\n","    def init_data_generator_parameters(self):\n","\n","        self.data_generator_parameters = dict(\n","            rescale=None,  # 1./255,\n","            preprocessing_function=tf.keras.applications.resnet50.preprocess_input\n","        )\n","\n","    def prepare_model(self):\n","        feature_extractor = resnet50.ResNet50(\n","            weights='imagenet',\n","            pooling='avg',\n","            input_shape=(224, 224, 3),\n","            include_top=False\n","        )\n","        x = feature_extractor.output\n","        x = Dense(units=400, activation='relu')(x)\n","        x = Dropout(0.3)(x)\n","        x = Dense(units=400, activation='relu')(x)\n","        x = Dropout(0.3)(x)\n","        outputs = Dense(units=1, activation='linear')(x)\n","\n","        self.model = Model(\n","            inputs=feature_extractor.input,\n","            outputs=outputs\n","        )"]},{"cell_type":"markdown","metadata":{"id":"oSRBPIeGdLGN"},"source":["## Xception"]},{"cell_type":"code","execution_count":20,"metadata":{"id":"IGwJgghhdS5I","executionInfo":{"status":"ok","timestamp":1707289472458,"user_tz":-60,"elapsed":6,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["from tensorflow.keras.applications.xception import Xception\n","\n","\n","class XceptionModel(NNModel):\n","    def __init__(self, *args, **kwargs):\n","        super().__init__(*args, **kwargs)\n","\n","    def init_data_generator_parameters(self):\n","\n","        self.data_generator_parameters = dict(\n","            rescale=None,  # 1./255,\n","            preprocessing_function=tf.keras.applications.xception.preprocess_input\n","        )\n","\n","    def prepare_model(self):\n","        feature_extractor = Xception(\n","            weights='imagenet',\n","            pooling='avg',\n","            input_shape=(299, 299, 3),\n","            include_top=False\n","        )\n","        x = feature_extractor.output\n","        x = Dense(units=400, activation='relu')(x)\n","        x = Dropout(0.3)(x)\n","        x = Dense(units=400, activation='relu')(x)\n","        x = Dropout(0.3)(x)\n","        outputs = Dense(units=1, activation='linear')(x)\n","\n","        self.model = Model(\n","            inputs=feature_extractor.input,\n","            outputs=outputs\n","        )"]},{"cell_type":"markdown","metadata":{"id":"K4DCM9x6dMSx"},"source":["## Inception V3"]},{"cell_type":"code","execution_count":21,"metadata":{"id":"zil9NH7tdXkD","executionInfo":{"status":"ok","timestamp":1707289472458,"user_tz":-60,"elapsed":6,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["from tensorflow.keras.applications.inception_v3 import InceptionV3\n","\n","\n","class InceptionV3Model(NNModel):\n","    def __init__(self, *args, **kwargs):\n","        super().__init__(*args, **kwargs)\n","\n","    def init_data_generator_parameters(self):\n","\n","        self.data_generator_parameters = dict(\n","            rescale=None,\n","            preprocessing_function=tf.keras.applications.inception_v3.preprocess_input\n","        )\n","\n","    def prepare_model(self):\n","        feature_extractor = InceptionV3(\n","            weights='imagenet',\n","            pooling='avg',\n","            input_shape=(299, 299, 3),\n","            include_top=False\n","        )\n","        x = feature_extractor.output\n","        x = Dense(units=400, activation='relu')(x)\n","        x = Dropout(0.3)(x)\n","        x = Dense(units=400, activation='relu')(x)\n","        x = Dropout(0.3)(x)\n","        outputs = Dense(units=1, activation='linear')(x)\n","\n","        self.model = Model(\n","            inputs=feature_extractor.input,\n","            outputs=outputs\n","        )"]},{"cell_type":"markdown","metadata":{"id":"hVwFlTXgdOQ4"},"source":["## Inception Resnet V2"]},{"cell_type":"code","execution_count":22,"metadata":{"id":"ilOK4s-Adav5","executionInfo":{"status":"ok","timestamp":1707289472458,"user_tz":-60,"elapsed":6,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["from tensorflow.keras.applications.inception_resnet_v2 import InceptionResNetV2\n","\n","\n","class InceptionResnetV2Model(NNModel):\n","    def __init__(self, *args, **kwargs):\n","        super().__init__(*args, **kwargs)\n","\n","    def init_data_generator_parameters(self):\n","\n","        self.data_generator_parameters = dict(\n","            rescale=None,\n","            preprocessing_function=tf.\\\n","                keras.\\\n","                applications.\\\n","                inception_resnet_v2.\\\n","                preprocess_input\n","        )\n","\n","    def prepare_model(self):\n","        feature_extractor = InceptionResNetV2(\n","            weights='imagenet',\n","            pooling='avg',\n","            input_shape=(299, 299, 3),\n","            include_top=False\n","        )\n","        x = feature_extractor.output\n","        x = Dense(units=400, activation='relu')(x)\n","        x = Dropout(0.3)(x)\n","        x = Dense(units=400, activation='relu')(x)\n","        x = Dropout(0.3)(x)\n","        outputs = Dense(units=1, activation='linear')(x)\n","\n","        self.model = Model(\n","            inputs=feature_extractor.input,\n","            outputs=outputs\n","        )"]},{"cell_type":"markdown","metadata":{"id":"gejwHXs7dQKq"},"source":["## VGG16"]},{"cell_type":"code","execution_count":23,"metadata":{"id":"2UY4ou5gdcyb","executionInfo":{"status":"ok","timestamp":1707289472458,"user_tz":-60,"elapsed":6,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["from tensorflow.keras.applications.vgg16 import VGG16\n","\n","\n","class VGG16Model(NNModel):\n","    def __init__(self, *args, **kwargs):\n","        super().__init__(*args, **kwargs)\n","\n","    def init_data_generator_parameters(self):\n","        print(f'VGG16: init data generator parameters')\n","        self.data_generator_parameters = dict(\n","            rescale=None,\n","            preprocessing_function=tf.keras.applications.vgg16.preprocess_input\n","        )\n","\n","    def prepare_model(self):\n","        feature_extractor = VGG16(\n","            weights='imagenet',\n","            pooling='avg',\n","            input_shape=(224, 224, 3),\n","            include_top=False\n","        )\n","        x = feature_extractor.output\n","        x = Dense(units=400, activation='relu')(x)\n","        x = Dropout(0.3)(x)\n","        x = Dense(units=400, activation='relu')(x)\n","        x = Dropout(0.3)(x)\n","        outputs = Dense(units=1, activation='linear')(x)\n","\n","        self.model = Model(\n","            inputs=feature_extractor.input,\n","            outputs=outputs\n","        )"]},{"cell_type":"markdown","metadata":{"id":"Xr4Bsi5LdRzl"},"source":["## VGG19"]},{"cell_type":"code","execution_count":24,"metadata":{"id":"RQvFRdnRdfIo","executionInfo":{"status":"ok","timestamp":1707289472458,"user_tz":-60,"elapsed":5,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[],"source":["from tensorflow.keras.applications.vgg19 import VGG19\n","\n","\n","class VGG19Model(NNModel):\n","    def __init__(self, *args, **kwargs):\n","        super().__init__(*args, **kwargs)\n","\n","    def init_data_generator_parameters(self):\n","        self.data_generator_parameters = dict(\n","            rescale=None,\n","            preprocessing_function=\\\n","                tf.keras.applications.vgg19.preprocess_input\n","        )\n","\n","    def prepare_model(self):\n","        feature_extractor = VGG19(\n","            weights='imagenet',\n","            pooling='avg',\n","            input_shape=(224, 224, 3),\n","            include_top=False\n","        )\n","        x = feature_extractor.output\n","        x = Dense(units=400, activation='relu')(x)\n","        x = Dropout(0.3)(x)\n","        x = Dense(units=400, activation='relu')(x)\n","        x = Dropout(0.3)(x)\n","        outputs = Dense(units=1, activation='linear')(x)\n","\n","        self.model = Model(\n","            inputs=feature_extractor.input,\n","            outputs=outputs\n","        )"]},{"cell_type":"markdown","metadata":{"id":"HcflYHf8doiH"},"source":["# Comparison of validation performance"]},{"cell_type":"code","execution_count":25,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":13305,"status":"ok","timestamp":1707289485758,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"},"user_tz":-60},"id":"411Hzvq9drQz","outputId":"575c13cc-e0f3-458e-fae3-10747d2a6275"},"outputs":[{"output_type":"stream","name":"stdout","text":["Found 1280 validated image filenames.\n","Found 320 validated image filenames.\n","Found 400 validated image filenames.\n","Found 1280 validated image filenames.\n","Found 320 validated image filenames.\n","Found 400 validated image filenames.\n","Found 1280 validated image filenames.\n","Found 320 validated image filenames.\n","Found 400 validated image filenames.\n","Found 1280 validated image filenames.\n","Found 320 validated image filenames.\n","Found 400 validated image filenames.\n","VGG16: init data generator parameters\n","Found 1280 validated image filenames.\n","Found 320 validated image filenames.\n","Found 400 validated image filenames.\n","Found 1280 validated image filenames.\n","Found 320 validated image filenames.\n","Found 400 validated image filenames.\n"]}],"source":["common_params = dict(\n","    validation_split_=0.2,\n","    batch_size_=64,\n","    trainval_labels_df_=trainval_labels_df,\n","    test_labels_df_=test_labels_df\n",")\n","\n","MODELS = {\n","    'resnet': ResNetModel(\n","        name_='resnet',\n","        input_image_size_=(224, 224),\n","        **common_params\n","    ),\n","    'xception': XceptionModel(\n","        name_='xception',\n","        input_image_size_=(299, 299),\n","        **common_params\n","    ),\n","    'inception_v3': InceptionV3Model(\n","        name_='inception_v3',\n","        input_image_size_=(299, 299),\n","        **common_params\n","    ),\n","    'inception_resnet_v2': InceptionResnetV2Model(\n","        name_='inception_resnet_v2',\n","        input_image_size_=(299, 299),\n","        **common_params\n","    ),\n","    'vgg16': VGG16Model(\n","        name_='vgg16',\n","        input_image_size_=(224, 224),\n","        **common_params\n","    ),\n","    'vgg19': VGG19Model(\n","        name_='vgg19',\n","        input_image_size_=(224, 224),\n","        **common_params\n","    )\n","}"]},{"cell_type":"code","execution_count":26,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"jlooHFcLf83B","outputId":"ee49d253-35b6-4793-9985-c27adeae4b3b","executionInfo":{"status":"ok","timestamp":1707289765252,"user_tz":-60,"elapsed":279500,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}}},"outputs":[{"output_type":"stream","name":"stdout","text":["Working with resnet\n","Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/resnet/resnet50_weights_tf_dim_ordering_tf_kernels_notop.h5\n","94765736/94765736 [==============================] - 1s 0us/step\n","5/5 [==============================] - 98s 22s/step - loss: 12.3364 - mse: 12.3364 - mae: 2.8113\n","Working with xception\n","Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/xception/xception_weights_tf_dim_ordering_tf_kernels_notop.h5\n","83683744/83683744 [==============================] - 0s 0us/step\n","5/5 [==============================] - 14s 446ms/step - loss: 11.2097 - mse: 11.2097 - mae: 2.6532\n","Working with inception_v3\n","Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/inception_v3/inception_v3_weights_tf_dim_ordering_tf_kernels_notop.h5\n","87910968/87910968 [==============================] - 1s 0us/step\n","5/5 [==============================] - 12s 312ms/step - loss: 15.6319 - mse: 15.6319 - mae: 3.1501\n","Working with inception_resnet_v2\n","Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/inception_resnet_v2/inception_resnet_v2_weights_tf_dim_ordering_tf_kernels_notop.h5\n","219055592/219055592 [==============================] - 1s 0us/step\n","5/5 [==============================] - 14s 612ms/step - loss: 31.6155 - mse: 31.6155 - mae: 4.4992\n","Working with vgg16\n","Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/vgg16/vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5\n","58889256/58889256 [==============================] - 1s 0us/step\n","5/5 [==============================] - 14s 298ms/step - loss: 13.3606 - mse: 13.3606 - mae: 2.9024\n","Working with vgg19\n","Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/vgg19/vgg19_weights_tf_dim_ordering_tf_kernels_notop.h5\n","80134624/80134624 [==============================] - 0s 0us/step\n","5/5 [==============================] - 3s 478ms/step - loss: 17.5642 - mse: 17.5642 - mae: 3.2782\n"]}],"source":["comparison_data = collect_comparison_data(MODELS)"]},{"cell_type":"code","execution_count":27,"metadata":{"id":"cActu2GOgQea","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1707289765253,"user_tz":-60,"elapsed":38,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}},"outputId":"2ab741fd-1708-4fb1-9e1e-e92ce60b4531"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["{'validation_results': {'resnet': [12.336389541625977,\n","   12.336389541625977,\n","   2.8113210201263428],\n","  'xception': [11.209732055664062, 11.209732055664062, 2.6532440185546875],\n","  'inception_v3': [15.631881713867188, 15.631881713867188, 3.1501400470733643],\n","  'inception_resnet_v2': [31.615463256835938,\n","   31.615463256835938,\n","   4.499157428741455],\n","  'vgg16': [13.360647201538086, 13.360647201538086, 2.9023914337158203],\n","  'vgg19': [17.56418800354004, 17.56418800354004, 3.278167247772217]},\n"," 'model_parameters': {'resnet': {'Total parameters': 24568113,\n","   'Trainable parameters': 24514993,\n","   'Non-trainable parameters': 53120},\n","  'xception': {'Total parameters': 21841881,\n","   'Trainable parameters': 21787353,\n","   'Non-trainable parameters': 54528},\n","  'inception_v3': {'Total parameters': 22783185,\n","   'Trainable parameters': 22748753,\n","   'Non-trainable parameters': 34432},\n","  'inception_resnet_v2': {'Total parameters': 55112337,\n","   'Trainable parameters': 55051793,\n","   'Non-trainable parameters': 60544},\n","  'vgg16': {'Total parameters': 15080689,\n","   'Trainable parameters': 15080689,\n","   'Non-trainable parameters': 0},\n","  'vgg19': {'Total parameters': 20390385,\n","   'Trainable parameters': 20390385,\n","   'Non-trainable parameters': 0}}}"]},"metadata":{},"execution_count":27}],"source":["comparison_data"]},{"cell_type":"code","execution_count":28,"metadata":{"id":"zLYIgmXPgotj","colab":{"base_uri":"https://localhost:8080/","height":737},"executionInfo":{"status":"ok","timestamp":1707289766041,"user_tz":-60,"elapsed":821,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}},"outputId":"aca4ff32-a383-46ba-f6c9-6cc5964c8f4d"},"outputs":[{"output_type":"stream","name":"stdout","text":["                            0\n","resnet               2.811321\n","xception             2.653244\n","inception_v3         3.150140\n","inception_resnet_v2  4.499157\n","vgg16                2.902391\n","vgg19                3.278167\n"]},{"output_type":"display_data","data":{"text/plain":["<Figure size 640x480 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}],"source":["plot_validation_results(comparison_data['validation_results'])"]},{"cell_type":"markdown","metadata":{"id":"RYTAmoESoxLX"},"source":["# Comparison of validation learning curves"]},{"cell_type":"code","execution_count":33,"metadata":{"id":"p6ub115eoy0S","colab":{"base_uri":"https://localhost:8080/","height":673},"executionInfo":{"status":"ok","timestamp":1707289771016,"user_tz":-60,"elapsed":6,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}},"outputId":"49debab3-5d69-4494-97d1-2449a2caa6a3"},"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 1000x700 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}],"source":["compare_learning_curve(\n","    {\n","        model_name: os.path.join(\n","            manager.models_trial_path,\n","            f'{model_object.name}.csv'\n","        )\n","        for model_name, model_object in MODELS.items()\n","    },\n","    metrics_='val_mae'\n",")"]},{"cell_type":"markdown","metadata":{"id":"59hdr7RH5NO1"},"source":["# Generalization ability VS number of unfreezed layers"]},{"cell_type":"code","execution_count":35,"metadata":{"id":"Y4BOk1Km5cS2","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1707289771827,"user_tz":-60,"elapsed":10,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}},"outputId":"4ce6863a-d33b-4878-8fff-e786ce5650be"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["{'xception_10_unfreezed_layers': 'xception_10_unfreezed_layers.csv',\n"," 'xception_15_unfreezed_layers': 'xception_15_unfreezed_layers.csv',\n"," 'xception_20_unfreezed_layers': 'xception_20_unfreezed_layers.csv',\n"," 'xception_25_unfreezed_layers': 'xception_25_unfreezed_layers.csv',\n"," 'xception_30_unfreezed_layers': 'xception_30_unfreezed_layers.csv',\n"," 'xception_40_unfreezed_layers': 'xception_40_unfreezed_layers.csv'}"]},"metadata":{},"execution_count":35}],"source":["filenames = [\n","    f'xception_{i}_unfreezed_layers' for i in [\n","        10, 15, 20, 25, 30, 40\n","    ]\n","]\n","csv_files_dict = {\n","    f: f'{f}.csv' for f in filenames\n","}\n","csv_files_dict"]},{"cell_type":"markdown","metadata":{"id":"ZkeE3pvn6sL-"},"source":["## Plotting validation performance"]},{"cell_type":"code","execution_count":38,"metadata":{"id":"i2HrwpsD8KlX","colab":{"base_uri":"https://localhost:8080/","height":504},"executionInfo":{"status":"ok","timestamp":1707289775182,"user_tz":-60,"elapsed":11,"user":{"displayName":"Süleyman Erim","userId":"14103164427425120523"}},"outputId":"2978dc83-a350-4540-e24d-156cad89256f"},"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 640x480 with 1 Axes>"],"image/png":"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\n"},"metadata":{}}],"source":["plot_validation_performance_vs_unfreezed_layers('val_mae')"]}],"metadata":{"accelerator":"GPU","colab":{"provenance":[{"file_id":"https://github.com/PashaIanko/Kaggle.CrowdCounting/blob/main/1_model_comparison.ipynb","timestamp":1705515109625}],"collapsed_sections":["neuR_JGoSki5","7qM_UgFLSqcF","Dr5iOQgfStF9","ZkKxmfS9SznR","EEhgFmlqS9at","fWzz3eYVTC9j","juSMgE8UTZ6B"]},"kernelspec":{"display_name":"Python 3","name":"python3"},"language_info":{"name":"python"}},"nbformat":4,"nbformat_minor":0}