src/Models.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hxOfDbrcvunw"
      },
      "source": [
        "# Model runs Notebook\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BiO3RUwhoQmG"
      },
      "source": [
        "In this notebook, various models are trained. Their predictions are then stored for evaluation. In order to obtain robust results, each version of the dataset (original samples only and then its copy combined with augmentation techniques) is split into 10 folds, allowing 10 unique combinations of train/test data to be used.\n",
        "\n",
        "In total, we have 11 versions of the dataset, with each being split into 10. We end up with at least 110 train+test runs for each model architecture, which creates a lot of computation demand. These tests were run in Google Colab. Results were then stored in a pickled dataframe. Prediction results were then analyzed in [Evaluation Notebook](Evaluation.ipynb)."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FWjLxxo7K9tH",
        "outputId": "74c2fa9d-1ec6-489a-a0ac-43837ed8dc6b"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: scikit-learn in /usr/local/lib/python3.10/dist-packages (1.2.2)\n",
            "Collecting scikit-learn\n",
            "  Downloading scikit_learn-1.3.2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (10.8 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m10.8/10.8 MB\u001b[0m \u001b[31m45.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: pandas in /usr/local/lib/python3.10/dist-packages (1.5.3)\n",
            "Collecting pandas\n",
            "  Downloading pandas-2.1.4-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (12.3 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m12.3/12.3 MB\u001b[0m \u001b[31m104.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: numpy<2.0,>=1.17.3 in /usr/local/lib/python3.10/dist-packages (from scikit-learn) (1.23.5)\n",
            "Requirement already satisfied: scipy>=1.5.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn) (1.11.4)\n",
            "Requirement already satisfied: joblib>=1.1.1 in /usr/local/lib/python3.10/dist-packages (from scikit-learn) (1.3.2)\n",
            "Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn) (3.2.0)\n",
            "Requirement already satisfied: python-dateutil>=2.8.2 in /usr/local/lib/python3.10/dist-packages (from pandas) (2.8.2)\n",
            "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas) (2023.3.post1)\n",
            "Collecting tzdata>=2022.1 (from pandas)\n",
            "  Downloading tzdata-2023.3-py2.py3-none-any.whl (341 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m341.8/341.8 kB\u001b[0m \u001b[31m42.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-packages (from python-dateutil>=2.8.2->pandas) (1.16.0)\n",
            "Installing collected packages: tzdata, scikit-learn, pandas\n",
            "  Attempting uninstall: scikit-learn\n",
            "    Found existing installation: scikit-learn 1.2.2\n",
            "    Uninstalling scikit-learn-1.2.2:\n",
            "      Successfully uninstalled scikit-learn-1.2.2\n",
            "  Attempting uninstall: pandas\n",
            "    Found existing installation: pandas 1.5.3\n",
            "    Uninstalling pandas-1.5.3:\n",
            "      Successfully uninstalled pandas-1.5.3\n",
            "\u001b[31mERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.\n",
            "lida 0.0.10 requires fastapi, which is not installed.\n",
            "lida 0.0.10 requires kaleido, which is not installed.\n",
            "lida 0.0.10 requires python-multipart, which is not installed.\n",
            "lida 0.0.10 requires uvicorn, which is not installed.\n",
            "google-colab 1.0.0 requires pandas==1.5.3, but you have pandas 2.1.4 which is incompatible.\u001b[0m\u001b[31m\n",
            "\u001b[0mSuccessfully installed pandas-2.1.4 scikit-learn-1.3.2 tzdata-2023.3\n"
          ]
        }
      ],
      "source": [
        "!pip install -U scikit-learn pandas\n",
        "# Original sklearn on Colab is 1.2.2 while 1.3.0 is used on my local machine"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kBi5sotNEqnK"
      },
      "source": [
        "### Step 1: Preparing results dataframe\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mfsv7JZyzK6H",
        "outputId": "0ad703c9-6568-455a-9b73-d2c513b5b57b"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Mounted at /content/gdrive\n"
          ]
        }
      ],
      "source": [
        "#This cell is all about setting results df, and major imports\n",
        "from google.colab import drive\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "path = '/content/gdrive/MyDrive/Colab Notebooks/data/'\n",
        "drive.mount('/content/gdrive')\n",
        "\n",
        "results = pd.read_pickle(\n",
        "    path + 'spectrum.pkl').reset_index()[['slice_file_name', 'fold', 'class']]\n",
        "augment_results = results.copy()\n",
        "results['source'] = 'o'\n",
        "augment_results['source'] = 'a'\n",
        "results = pd.concat([results, augment_results], ignore_index=True, axis=0)\n",
        "results.to_pickle(path + \"results.pkl\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2IlrBm8NzoTS"
      },
      "source": [
        "### Main Execution Part\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YNpIGRChYAYQ",
        "outputId": "0c2b0f0a-96ab-49e6-91bb-91053f8d2d6d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/gdrive; to attempt to forcibly remount, call drive.mount(\"/content/gdrive\", force_remount=True).\n"
          ]
        }
      ],
      "source": [
        "from sklearn.preprocessing import OneHotEncoder\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from google.colab import drive\n",
        "path = '/content/gdrive/MyDrive/Colab Notebooks/data/'\n",
        "drive.mount('/content/gdrive')\n",
        "names = ['dist', 'mixup', 'imixup', 'room', 'warp', 'delay', 'spectrum']\n",
        "classes = ['air_conditioner', 'car_horn', 'children_playing', 'dog_bark',\n",
        "           'drilling', 'engine_idling', 'gun_shot', 'jackhammer', 'siren', 'street_music']\n",
        "num_classes = len(classes)\n",
        "\n",
        "# This parameter is specifying what kind of model is going to be used for the experiment run.\n",
        "backbone = \"RsNet50\"\n",
        "\n",
        "\n",
        "def build_df(augment_name):\n",
        "    \"\"\"\n",
        "    This function is effectively building a working dataframe from unaugmented data and selected (augment_name) augmented dataframe.\n",
        "    One-hot encoding is also done here.\n",
        "    \"\"\"\n",
        "    original_df = pd.read_pickle(path + 'spectrum.pkl')\n",
        "    aug_df = pd.read_pickle(path + augment_name + '.pkl')\n",
        "    enc = OneHotEncoder(handle_unknown='ignore',\n",
        "                        feature_name_combiner=lambda x, y: y)\n",
        "    categories_df = pd.DataFrame(enc.fit_transform(original_df['class'].to_numpy(\n",
        "    ).reshape(-1, 1)).toarray(), columns=enc.get_feature_names_out())\n",
        "    categories_df = pd.concat(\n",
        "        [original_df[['slice_file_name', 'fold', 'r']], categories_df], axis=1)\n",
        "    categories_df['source'] = 'o'\n",
        "    if augment_name != 'spectrum':\n",
        "        if augment_name == 'mixup':\n",
        "            encoding = aug_df.set_index('classID').to_dict()['class']\n",
        "            aug_df['class2'] = aug_df['classID2'].map(encoding)\n",
        "            categories2_df = pd.DataFrame(enc.transform(aug_df['class'].to_numpy(\n",
        "            ).reshape(-1, 1)).toarray(), columns=enc.get_feature_names_out())*0.8\n",
        "            categories2_df_add = pd.DataFrame(enc.transform(aug_df['class2'].to_numpy(\n",
        "            ).reshape(-1, 1)).toarray(), columns=enc.get_feature_names_out())*0.2\n",
        "            categories2_df = categories2_df.add(categories2_df_add)\n",
        "            categories2_df = pd.concat(\n",
        "                [aug_df[['slice_file_name', 'fold', 'r']], categories2_df], axis=1)\n",
        "        elif augment_name == 'imixup':\n",
        "            encoding = aug_df.set_index('classID').to_dict()['class']\n",
        "            aug_df['class2'] = aug_df['classID2'].map(encoding)\n",
        "            categories2_df = pd.DataFrame(enc.transform(aug_df['class'].to_numpy(\n",
        "            ).reshape(-1, 1)).toarray(), columns=enc.get_feature_names_out())*0.66\n",
        "            categories2_df_add = pd.DataFrame(enc.transform(aug_df['class2'].to_numpy(\n",
        "            ).reshape(-1, 1)).toarray(), columns=enc.get_feature_names_out())*0.34\n",
        "            categories2_df = categories2_df.add(categories2_df_add)\n",
        "            categories2_df = pd.concat(\n",
        "                [aug_df[['slice_file_name', 'fold', 'r']], categories2_df], axis=1)\n",
        "        else:\n",
        "            categories2_df = pd.DataFrame(enc.transform(aug_df['class'].to_numpy(\n",
        "            ).reshape(-1, 1)).toarray(), columns=enc.get_feature_names_out())\n",
        "            categories2_df = pd.concat(\n",
        "                [aug_df[['slice_file_name', 'fold', 'r']], categories2_df], axis=1)\n",
        "        categories2_df['source'] = 'a'\n",
        "        return [categories_df, categories2_df]\n",
        "    else:\n",
        "        return [categories_df]\n",
        "\n",
        "\n",
        "def split(df, fold):\n",
        "    \"\"\"\n",
        "    This function is splitting previously built dataframe into train/test/val x/y arrays. Based on folds predefined in a dataset\n",
        "    \"\"\"\n",
        "    train = df[df['fold'] != fold]\n",
        "    val_fold = train.fold.sample(n=1).values[0]\n",
        "    val = train[train['fold'] == val_fold]\n",
        "    train = train[train['fold'] != val_fold]\n",
        "    test = df[df['fold'] == fold]\n",
        "    train_x = np.stack([x for x in train['r'].values])\n",
        "    train_y = train[classes]\n",
        "    test = df[df['fold'] == fold]\n",
        "    test_x = np.stack([x for x in test['r'].values])\n",
        "    test_y = test[classes]\n",
        "    val_x = np.stack([x for x in val['r'].values])\n",
        "    val_y = val[classes]\n",
        "    # print(len(val))\n",
        "    return train_x, train_y, test_x, test_y, val_x, val_y"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "d-zwaIuFEqnN"
      },
      "outputs": [],
      "source": [
        "import tensorflow as tf\n",
        "import tensorflow_hub as hub\n",
        "from tensorflow.keras.layers import Input, Dense, Dropout, SpatialDropout1D, GlobalAveragePooling1D, GlobalMaxPooling1D\n",
        "from tensorflow.keras.models import Model\n",
        "from tensorflow.keras.utils import plot_model\n",
        "import time\n",
        "import os\n",
        "\n",
        "def file_control(fpath):\n",
        "    \"\"\"\n",
        "    This function is used to clear the file storing weights of the model. If not done,\n",
        "    ModelCheckpoint will capture best performance from the file, ignoring consequent experiment runs.\n",
        "    \"\"\"\n",
        "    if os.path.isfile(fpath):\n",
        "        open(fpath, 'w').close()\n",
        "        os.remove(fpath)\n",
        "        print('cleared old weights')\n",
        "    else:\n",
        "        print('already clear')\n",
        "\n",
        "\n",
        "def build_model(nc=1, shape=(224, 224, 3), base_trainable=False, backbone = \"EfNetV2B1\", weights = None):\n",
        "    \"\"\"\n",
        "    This is a function creating a {backbone}-based model. The base model itself is pulled from tfhub/tf repo, Following paraneters are to be specified when calling this function:\n",
        "    nc -- number of classes, or \"width\" of the model head\n",
        "    shape -- image size provided for an input\n",
        "    base_trainable -- layers of the base model (EfficientNetV2B2 in this case) can be locked as trainable or not, affecting model performance and training time.\n",
        "    backbone - type of backbone to use\n",
        "    weights - None or 'imagenet'\n",
        "    \"\"\"\n",
        "    if backbone == \"EfNetV2B1\":\n",
        "      base_model = tf.keras.applications.efficientnet_v2.EfficientNetV2B1(include_top=False, input_shape=shape, pooling='max', weights=weights)\n",
        "    elif backbone == \"EfNetV2B2\":\n",
        "      base_model=tf.keras.applications.efficientnet_v2.EfficientNetV2B2(include_top=False, input_shape=shape, pooling='max', weights=weights)\n",
        "    elif backbone == \"ResNet152\":\n",
        "      base_model=tf.keras.applications.resnet_v2.ResNet152V2(include_top=False, input_shape=shape, pooling='max', weights=weights)\n",
        "    else:\n",
        "      base_model=tf.keras.applications.resnet_v2.ResNet50V2(include_top=False, input_shape=shape, pooling='max', weights=weights)\n",
        "    for layer in base_model.layers:\n",
        "        layer.trainable = base_trainable\n",
        "    dropout_layer = tf.keras.layers.Dropout(\n",
        "        0.5, name=\"Dropout1\")(base_model.output)\n",
        "    clf_layer = tf.keras.layers.Dense(\n",
        "        256, activation=\"sigmoid\", name='Dense1')(dropout_layer)\n",
        "    dropout_layer2 = tf.keras.layers.Dropout(0.4, name=\"Dropout2\")(clf_layer)\n",
        "    clf_layer2 = tf.keras.layers.Dense(\n",
        "        128, activation=\"sigmoid\", name='Dense2')(dropout_layer2)\n",
        "    dropout_layer3 = tf.keras.layers.Dropout(0.1, name=\"Dropout3\")(clf_layer2)\n",
        "    clf_layer3 = tf.keras.layers.Dense(\n",
        "        nc, activation=\"sigmoid\", name='Dense3')(dropout_layer3)\n",
        "    model = tf.keras.Model(base_model.input, clf_layer3)\n",
        "    model.compile(loss='categorical_crossentropy',\n",
        "                  optimizer='adam', metrics=[tf.keras.metrics.AUC()])\n",
        "    return model\n",
        "\n",
        "\n",
        "def create_dataset(x, y, batch_size=64, shuffle=True):\n",
        "  \"\"\" This function is converting pandas dataframe into tensorflow dataset.\n",
        "  This step should allow dynamic dataset loading into GPU RAM.\n",
        "  \"\"\"\n",
        "    dataset = tf.data.Dataset.from_tensor_slices((x, y))\n",
        "    if shuffle:\n",
        "        dataset = dataset.shuffle(buffer_size=len(x))\n",
        "    dataset = dataset.batch(batch_size)\n",
        "    return dataset\n",
        "\n",
        "\n",
        "def run_model(train_x, train_y, test_x, test_y, val_x, val_y, backbone = \"ResNet50\", batch_size = 64):\n",
        "    \"\"\" A Wrap-up function to run the model. It will be called from the main function in the next cell\n",
        "    Basically, it is creating a model, training it, and then returning the predictions on the test set.\n",
        "    \"\"\"\n",
        "    file_control(path + 'm.h5')\n",
        "    image_shape = np.shape(train_x[0])\n",
        "    num_classes = np.shape(train_y)[1]\n",
        "    # All models are given a chance to train for up to 100 epochs. However, in order to prevent overfit,\n",
        "    # Models will be stopped if their validation metric is not imporving for at least 1% for over 8 epochs\n",
        "    sb = tf.keras.callbacks.ModelCheckpoint(\n",
        "        path+'m.h5', monitor='val_loss', verbose=1, save_best_only=True)\n",
        "    es = tf.keras.callbacks.EarlyStopping(\n",
        "        monitor='val_loss', min_delta=0.01, patience=8, restore_best_weights=False)\n",
        "    train_dataset = create_dataset(train_x, train_y, batch_size = batch_size)\n",
        "    val_dataset = create_dataset(val_x, val_y, shuffle=False, batch_size = batch_size)\n",
        "    model = build_model(nc=num_classes, base_trainable=True, shape=image_shape, backbone = backbone)\n",
        "    model.fit(train_x, train_y, batch_size=64, epochs=100, verbose=0,\n",
        "              callbacks=[es, sb], validation_data=(val_x, val_y))\n",
        "    model.load_weights(path + 'm.h5')\n",
        "    test_dataset = create_dataset(test_x, None, shuffle=False,batch_size = batch_size)\n",
        "    predictions = model.predict(test_dataset)\n",
        "    return predictions"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "c0ahKxWMZw4l"
      },
      "outputs": [],
      "source": [
        "def run_experiments(name, backbone = \"RsNet50\", batch_size = 64):\n",
        "    \"\"\"\n",
        "    In this function, we are running the model for each fold of the dataset. The results are stored in the results.pkl file.\n",
        "    name is specyfing the version of the dataset to run experiments on.\n",
        "    \"\"\"\n",
        "    start = time.time()\n",
        "    print(f'running experiment for {name}')\n",
        "    results = pd.read_pickle(path + 'results.pkl')\n",
        "    df = pd.concat(build_df(name), ignore_index=True, axis=0)\n",
        "    r = pd.DataFrame(columns=['slice_file_name', 'fold', 'source', name])\n",
        "    for fold in results.fold.unique():\n",
        "        train_x, train_y, test_x, test_y, val_x, val_y = split(df, fold)\n",
        "        test_link = df[df['fold'] == fold][[\n",
        "            'slice_file_name', 'fold', 'source']]\n",
        "        pred = run_model(train_x, train_y, test_x, test_y, val_x, val_y, backbone = backbone)\n",
        "        test_link[name] = pd.DataFrame(\n",
        "            pred, columns=classes).idxmax(axis=1).to_numpy()\n",
        "        r = pd.concat(\n",
        "            [r, test_link[['slice_file_name', 'fold', 'source', name]]], ignore_index=True)\n",
        "    results = pd.merge(results, r,  how='left', left_on=[\n",
        "                       'slice_file_name', 'fold', 'source'], right_on=['slice_file_name', 'fold', 'source'])\n",
        "    results.to_pickle(path + \"results.pkl\")\n",
        "    end = time.time()\n",
        "    print(f'done in {str(round(end - start,2))}')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "a8s7tctJ1nTa",
        "outputId": "eaad4b68-89aa-41cb-d2d2-b11bd908e099"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "running experiment for dist\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.37302, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.37302\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.37302\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.37302\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.37302\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.37302\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.37302\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.37302\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.37302\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.72819, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.72819\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.72819\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.72819\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.72819\n",
            "\n",
            "Epoch 6: val_loss improved from 2.72819 to 1.91724, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss did not improve from 1.91724\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.91724\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.91724\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.91724\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.91724\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.91724\n",
            "\n",
            "Epoch 13: val_loss improved from 1.91724 to 1.83417, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.83417\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.83417\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.83417\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.83417\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.83417\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.83417\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.83417\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.83417\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.59412, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.59412\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.59412\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.59412\n",
            "\n",
            "Epoch 5: val_loss improved from 2.59412 to 2.50050, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.50050\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.50050\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.50050\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.50050\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.50050\n",
            "\n",
            "Epoch 11: val_loss improved from 2.50050 to 2.37367, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 12: val_loss improved from 2.37367 to 2.08357, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 13: val_loss did not improve from 2.08357\n",
            "\n",
            "Epoch 14: val_loss did not improve from 2.08357\n",
            "\n",
            "Epoch 15: val_loss did not improve from 2.08357\n",
            "\n",
            "Epoch 16: val_loss did not improve from 2.08357\n",
            "\n",
            "Epoch 17: val_loss did not improve from 2.08357\n",
            "\n",
            "Epoch 18: val_loss improved from 2.08357 to 1.95293, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.95293\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.95293\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.95293\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.95293\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.95293\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.95293\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.95293\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.95293\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.52860, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.52860\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.52860\n",
            "\n",
            "Epoch 4: val_loss improved from 2.52860 to 2.43663, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.43663\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.43663\n",
            "\n",
            "Epoch 7: val_loss improved from 2.43663 to 1.88126, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.88126\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.88126\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.88126\n",
            "\n",
            "Epoch 11: val_loss improved from 1.88126 to 1.41531, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.41531\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.41531\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.41531\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.41531\n",
            "\n",
            "Epoch 16: val_loss improved from 1.41531 to 1.31784, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.31784\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.31784\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.31784\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.31784\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.31784\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.31784\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.31784\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.31784\n",
            "6/6 [==============================] - 2s 171ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.41902, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.41902\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.41902\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.41902\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.41902\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.41902\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.41902\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.41902\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.41902\n",
            "6/6 [==============================] - 1s 57ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.38619, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.38619\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.38619\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.38619\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.38619\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.38619\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.38619\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.38619\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.38619\n",
            "6/6 [==============================] - 2s 156ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.41936, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.41936\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.41936\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.41936\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.41936\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.41936\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.41936\n",
            "\n",
            "Epoch 8: val_loss improved from 2.41936 to 2.40801, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.40801\n",
            "\n",
            "Epoch 10: val_loss improved from 2.40801 to 2.09726, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.09726\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.09726\n",
            "\n",
            "Epoch 13: val_loss did not improve from 2.09726\n",
            "\n",
            "Epoch 14: val_loss did not improve from 2.09726\n",
            "\n",
            "Epoch 15: val_loss did not improve from 2.09726\n",
            "\n",
            "Epoch 16: val_loss did not improve from 2.09726\n",
            "\n",
            "Epoch 17: val_loss did not improve from 2.09726\n",
            "\n",
            "Epoch 18: val_loss did not improve from 2.09726\n",
            "6/6 [==============================] - 1s 57ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.41803, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.41803\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.41803\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.41803\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.41803\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.41803\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.41803\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.41803\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.41803\n",
            "6/6 [==============================] - 2s 145ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.43848, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.43848\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.43848\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.43848\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.43848\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.43848\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.43848\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.43848\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.43848\n",
            "6/6 [==============================] - 1s 57ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.51621, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.51621\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.51621\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.51621\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.51621\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.51621\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.51621\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.51621\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.51621\n",
            "6/6 [==============================] - 39s 56ms/step\n",
            "done in 1759.19\n"
          ]
        }
      ],
      "source": [
        "run_experiments(names[0], backbone = backbone)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FoCprjPE11PQ",
        "outputId": "d5cf882f-316d-45d4-99f7-4b0019fd7521"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "running experiment for mixup\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.48083, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.48083\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.48083\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.48083\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.48083\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.48083\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.48083\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.48083\n",
            "\n",
            "Epoch 9: val_loss improved from 2.48083 to 1.99550, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.99550\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.99550\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.99550\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.99550\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.99550\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.99550\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.99550\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.99550\n",
            "6/6 [==============================] - 1s 57ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.64853, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.64853\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.64853\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.64853\n",
            "\n",
            "Epoch 5: val_loss improved from 2.64853 to 2.47025, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.47025\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.47025\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.47025\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.47025\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.47025\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.47025\n",
            "\n",
            "Epoch 12: val_loss improved from 2.47025 to 2.43407, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 13: val_loss improved from 2.43407 to 2.17018, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 14: val_loss did not improve from 2.17018\n",
            "\n",
            "Epoch 15: val_loss did not improve from 2.17018\n",
            "\n",
            "Epoch 16: val_loss did not improve from 2.17018\n",
            "\n",
            "Epoch 17: val_loss did not improve from 2.17018\n",
            "\n",
            "Epoch 18: val_loss did not improve from 2.17018\n",
            "\n",
            "Epoch 19: val_loss improved from 2.17018 to 1.89268, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.89268\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.89268\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.89268\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.89268\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.89268\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.89268\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.89268\n",
            "\n",
            "Epoch 27: val_loss did not improve from 1.89268\n",
            "6/6 [==============================] - 1s 57ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.53945, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.53945\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.53945\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.53945\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.53945\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.53945\n",
            "\n",
            "Epoch 7: val_loss improved from 2.53945 to 2.21400, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.21400\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.21400\n",
            "\n",
            "Epoch 10: val_loss improved from 2.21400 to 2.21167, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss improved from 2.21167 to 1.83240, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.83240\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.83240\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.83240\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.83240\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.83240\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.83240\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.83240\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.83240\n",
            "6/6 [==============================] - 3s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.33529, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.33529\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.33529\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.33529\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.33529\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.33529\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.33529\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.33529\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.33529\n",
            "6/6 [==============================] - 1s 55ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.59334, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.59334\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.59334\n",
            "\n",
            "Epoch 4: val_loss improved from 2.59334 to 2.38577, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.38577\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.38577\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.38577\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.38577\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.38577\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.38577\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.38577\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.38577\n",
            "6/6 [==============================] - 1s 55ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.42508, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.42508\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.42508\n",
            "\n",
            "Epoch 4: val_loss improved from 2.42508 to 2.38165, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.38165\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.38165\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.38165\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.38165\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.38165\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.38165\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.38165\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.38165\n",
            "6/6 [==============================] - 1s 57ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.64652, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.64652\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.64652\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.64652\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.64652\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.64652\n",
            "\n",
            "Epoch 7: val_loss improved from 2.64652 to 2.39776, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.39776\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.39776\n",
            "\n",
            "Epoch 10: val_loss improved from 2.39776 to 2.08130, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.08130\n",
            "\n",
            "Epoch 12: val_loss improved from 2.08130 to 1.99883, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.99883\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.99883\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.99883\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.99883\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.99883\n",
            "\n",
            "Epoch 18: val_loss improved from 1.99883 to 1.91151, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.91151\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.91151\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.91151\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.91151\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.91151\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.91151\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.91151\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.91151\n",
            "6/6 [==============================] - 1s 58ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.43921, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.43921\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.43921\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.43921\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.43921\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.43921\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.43921\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.43921\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.43921\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.44203, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.44203\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.44203\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.44203\n",
            "\n",
            "Epoch 5: val_loss improved from 2.44203 to 2.43111, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.43111\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.43111\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.43111\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.43111\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.43111\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.43111\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.43111\n",
            "\n",
            "Epoch 13: val_loss did not improve from 2.43111\n",
            "6/6 [==============================] - 1s 57ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.45218, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.45218\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.45218\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.45218\n",
            "\n",
            "Epoch 5: val_loss improved from 2.45218 to 2.40241, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss improved from 2.40241 to 2.29496, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.29496\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.29496\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.29496\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.29496\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.29496\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.29496\n",
            "\n",
            "Epoch 13: val_loss did not improve from 2.29496\n",
            "\n",
            "Epoch 14: val_loss improved from 2.29496 to 2.18920, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 15: val_loss did not improve from 2.18920\n",
            "\n",
            "Epoch 16: val_loss did not improve from 2.18920\n",
            "\n",
            "Epoch 17: val_loss did not improve from 2.18920\n",
            "\n",
            "Epoch 18: val_loss did not improve from 2.18920\n",
            "\n",
            "Epoch 19: val_loss did not improve from 2.18920\n",
            "\n",
            "Epoch 20: val_loss did not improve from 2.18920\n",
            "\n",
            "Epoch 21: val_loss did not improve from 2.18920\n",
            "\n",
            "Epoch 22: val_loss improved from 2.18920 to 2.04389, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 23: val_loss did not improve from 2.04389\n",
            "\n",
            "Epoch 24: val_loss did not improve from 2.04389\n",
            "\n",
            "Epoch 25: val_loss did not improve from 2.04389\n",
            "\n",
            "Epoch 26: val_loss did not improve from 2.04389\n",
            "\n",
            "Epoch 27: val_loss did not improve from 2.04389\n",
            "\n",
            "Epoch 28: val_loss did not improve from 2.04389\n",
            "\n",
            "Epoch 29: val_loss did not improve from 2.04389\n",
            "\n",
            "Epoch 30: val_loss did not improve from 2.04389\n",
            "6/6 [==============================] - 1s 57ms/step\n",
            "done in 2005.43\n"
          ]
        }
      ],
      "source": [
        "run_experiments(names[1],backbone = backbone)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tdrBuJC912UW",
        "outputId": "979babee-844d-421d-cf76-5e7fbd8c6c9a"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "running experiment for imixup\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.38907, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.38907\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.38907\n",
            "\n",
            "Epoch 4: val_loss improved from 2.38907 to 2.35374, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.35374\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.35374\n",
            "\n",
            "Epoch 7: val_loss improved from 2.35374 to 2.27097, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.27097\n",
            "\n",
            "Epoch 9: val_loss improved from 2.27097 to 2.20245, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.20245\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.20245\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.20245\n",
            "\n",
            "Epoch 13: val_loss did not improve from 2.20245\n",
            "\n",
            "Epoch 14: val_loss did not improve from 2.20245\n",
            "\n",
            "Epoch 15: val_loss did not improve from 2.20245\n",
            "\n",
            "Epoch 16: val_loss did not improve from 2.20245\n",
            "\n",
            "Epoch 17: val_loss improved from 2.20245 to 1.65332, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.65332\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.65332\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.65332\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.65332\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.65332\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.65332\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.65332\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.65332\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.37013, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.37013\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.37013\n",
            "\n",
            "Epoch 4: val_loss improved from 2.37013 to 2.33932, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.33932\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.33932\n",
            "\n",
            "Epoch 7: val_loss improved from 2.33932 to 1.94374, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.94374\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.94374\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.94374\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.94374\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.94374\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.94374\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.94374\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.94374\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.36535, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.36535\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.36535\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.36535\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.36535\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.36535\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.36535\n",
            "\n",
            "Epoch 8: val_loss improved from 2.36535 to 2.18217, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.18217\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.18217\n",
            "\n",
            "Epoch 11: val_loss improved from 2.18217 to 1.96233, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.96233\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.96233\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.96233\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.96233\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.96233\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.96233\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.96233\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.96233\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.32457, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.32457\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.32457\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.32457\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.32457\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.32457\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.32457\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.32457\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.32457\n",
            "6/6 [==============================] - 2s 171ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.42558, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.42558\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.42558\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.42558\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.42558\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.42558\n",
            "\n",
            "Epoch 7: val_loss improved from 2.42558 to 2.41269, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.41269\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.41269\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.41269\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.41269\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.41269\n",
            "\n",
            "Epoch 13: val_loss did not improve from 2.41269\n",
            "\n",
            "Epoch 14: val_loss improved from 2.41269 to 1.73948, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.73948\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.73948\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.73948\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.73948\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.73948\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.73948\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.73948\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.73948\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.50588, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.50588\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.50588\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.50588\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.50588\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.50588\n",
            "\n",
            "Epoch 7: val_loss improved from 2.50588 to 2.26368, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.26368\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.26368\n",
            "\n",
            "Epoch 10: val_loss improved from 2.26368 to 1.65207, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.65207\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.65207\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.65207\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.65207\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.65207\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.65207\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.65207\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.65207\n",
            "6/6 [==============================] - 2s 151ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.47855, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.47855\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.47855\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.47855\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.47855\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.47855\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.47855\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.47855\n",
            "\n",
            "Epoch 9: val_loss improved from 2.47855 to 2.23673, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.23673\n",
            "\n",
            "Epoch 11: val_loss improved from 2.23673 to 2.04508, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.04508\n",
            "\n",
            "Epoch 13: val_loss did not improve from 2.04508\n",
            "\n",
            "Epoch 14: val_loss did not improve from 2.04508\n",
            "\n",
            "Epoch 15: val_loss did not improve from 2.04508\n",
            "\n",
            "Epoch 16: val_loss did not improve from 2.04508\n",
            "\n",
            "Epoch 17: val_loss did not improve from 2.04508\n",
            "\n",
            "Epoch 18: val_loss did not improve from 2.04508\n",
            "\n",
            "Epoch 19: val_loss did not improve from 2.04508\n",
            "6/6 [==============================] - 1s 57ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.38191, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.38191\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.38191\n",
            "\n",
            "Epoch 4: val_loss improved from 2.38191 to 2.32801, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.32801\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.32801\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.32801\n",
            "\n",
            "Epoch 8: val_loss improved from 2.32801 to 2.25685, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.25685\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.25685\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.25685\n",
            "\n",
            "Epoch 12: val_loss improved from 2.25685 to 2.19391, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 13: val_loss did not improve from 2.19391\n",
            "\n",
            "Epoch 14: val_loss did not improve from 2.19391\n",
            "\n",
            "Epoch 15: val_loss did not improve from 2.19391\n",
            "\n",
            "Epoch 16: val_loss improved from 2.19391 to 1.95272, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.95272\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.95272\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.95272\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.95272\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.95272\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.95272\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.95272\n",
            "\n",
            "Epoch 24: val_loss improved from 1.95272 to 1.92543, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.92543\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.92543\n",
            "\n",
            "Epoch 27: val_loss improved from 1.92543 to 1.76266, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 28: val_loss did not improve from 1.76266\n",
            "\n",
            "Epoch 29: val_loss did not improve from 1.76266\n",
            "\n",
            "Epoch 30: val_loss did not improve from 1.76266\n",
            "\n",
            "Epoch 31: val_loss did not improve from 1.76266\n",
            "\n",
            "Epoch 32: val_loss did not improve from 1.76266\n",
            "\n",
            "Epoch 33: val_loss did not improve from 1.76266\n",
            "\n",
            "Epoch 34: val_loss did not improve from 1.76266\n",
            "\n",
            "Epoch 35: val_loss did not improve from 1.76266\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.38676, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.38676\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.38676\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.38676\n",
            "\n",
            "Epoch 5: val_loss improved from 2.38676 to 2.27408, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.27408\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.27408\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.27408\n",
            "\n",
            "Epoch 9: val_loss improved from 2.27408 to 1.97535, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.97535\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.97535\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.97535\n",
            "\n",
            "Epoch 13: val_loss improved from 1.97535 to 1.76363, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.76363\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.76363\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.76363\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.76363\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.76363\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.76363\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.76363\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.76363\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.43148, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.43148\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.43148\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.43148\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.43148\n",
            "\n",
            "Epoch 6: val_loss improved from 2.43148 to 2.15908, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.15908\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.15908\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.15908\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.15908\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.15908\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.15908\n",
            "\n",
            "Epoch 13: val_loss did not improve from 2.15908\n",
            "\n",
            "Epoch 14: val_loss did not improve from 2.15908\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "done in 2301.25\n"
          ]
        }
      ],
      "source": [
        "run_experiments(names[2],backbone = backbone)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9B5roDUw13LG",
        "outputId": "5738edea-5060-4eb7-8240-2a2f72be57c8"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "running experiment for room\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.45539, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.45539\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.45539\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.45539\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.45539\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.45539\n",
            "\n",
            "Epoch 7: val_loss improved from 2.45539 to 2.42564, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.42564\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.42564\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.42564\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.42564\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.42564\n",
            "\n",
            "Epoch 13: val_loss improved from 2.42564 to 2.25795, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 14: val_loss did not improve from 2.25795\n",
            "\n",
            "Epoch 15: val_loss did not improve from 2.25795\n",
            "\n",
            "Epoch 16: val_loss did not improve from 2.25795\n",
            "\n",
            "Epoch 17: val_loss did not improve from 2.25795\n",
            "\n",
            "Epoch 18: val_loss did not improve from 2.25795\n",
            "\n",
            "Epoch 19: val_loss improved from 2.25795 to 1.62881, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.62881\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.62881\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.62881\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.62881\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.62881\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.62881\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.62881\n",
            "\n",
            "Epoch 27: val_loss did not improve from 1.62881\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.36070, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.36070\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.36070\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.36070\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.36070\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.36070\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.36070\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.36070\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.36070\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.58896, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.58896\n",
            "\n",
            "Epoch 3: val_loss improved from 2.58896 to 2.56003, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss improved from 2.56003 to 2.50983, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.50983\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.50983\n",
            "\n",
            "Epoch 7: val_loss improved from 2.50983 to 2.38619, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss improved from 2.38619 to 2.31379, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.31379\n",
            "\n",
            "Epoch 10: val_loss improved from 2.31379 to 2.18338, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.18338\n",
            "\n",
            "Epoch 12: val_loss improved from 2.18338 to 1.69462, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.69462\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.69462\n",
            "\n",
            "Epoch 15: val_loss improved from 1.69462 to 1.56081, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.56081\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.56081\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.56081\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.56081\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.56081\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.56081\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.56081\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.56081\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.38228, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.38228\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.38228\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.38228\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.38228\n",
            "\n",
            "Epoch 6: val_loss improved from 2.38228 to 2.15749, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.15749\n",
            "\n",
            "Epoch 8: val_loss improved from 2.15749 to 1.96105, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.96105\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.96105\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.96105\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.96105\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.96105\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.96105\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.96105\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.96105\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.59246, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.59246\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.59246\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.59246\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.59246\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.59246\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.59246\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.59246\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.59246\n",
            "6/6 [==============================] - 1s 57ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.52498, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.52498\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.52498\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.52498\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.52498\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.52498\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.52498\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.52498\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.52498\n",
            "6/6 [==============================] - 1s 55ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.51579, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.51579\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.51579\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.51579\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.51579\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.51579\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.51579\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.51579\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.51579\n",
            "6/6 [==============================] - 1s 55ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.55090, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.55090\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.55090\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.55090\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.55090\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.55090\n",
            "\n",
            "Epoch 7: val_loss improved from 2.55090 to 2.49183, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss improved from 2.49183 to 2.45484, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.45484\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.45484\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.45484\n",
            "\n",
            "Epoch 12: val_loss improved from 2.45484 to 2.14505, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 13: val_loss improved from 2.14505 to 2.05163, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 14: val_loss improved from 2.05163 to 1.63781, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.63781\n",
            "\n",
            "Epoch 16: val_loss improved from 1.63781 to 1.43079, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.43079\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.43079\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.43079\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.43079\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.43079\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.43079\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.43079\n",
            "\n",
            "Epoch 24: val_loss improved from 1.43079 to 1.25550, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.25550\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.25550\n",
            "\n",
            "Epoch 27: val_loss did not improve from 1.25550\n",
            "\n",
            "Epoch 28: val_loss did not improve from 1.25550\n",
            "\n",
            "Epoch 29: val_loss did not improve from 1.25550\n",
            "\n",
            "Epoch 30: val_loss did not improve from 1.25550\n",
            "\n",
            "Epoch 31: val_loss did not improve from 1.25550\n",
            "\n",
            "Epoch 32: val_loss did not improve from 1.25550\n",
            "6/6 [==============================] - 1s 55ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.39393, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.39393\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.39393\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.39393\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.39393\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.39393\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.39393\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.39393\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.39393\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.59867, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.59867\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.59867\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.59867\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.59867\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.59867\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.59867\n",
            "\n",
            "Epoch 8: val_loss improved from 2.59867 to 2.23944, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.23944\n",
            "\n",
            "Epoch 10: val_loss improved from 2.23944 to 2.11637, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.11637\n",
            "\n",
            "Epoch 12: val_loss improved from 2.11637 to 1.63988, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.63988\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.63988\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.63988\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.63988\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.63988\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.63988\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.63988\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.63988\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "done in 1979.31\n"
          ]
        }
      ],
      "source": [
        "run_experiments(names[3],backbone = backbone)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fZ3VF1MG12pN",
        "outputId": "2f2b62b9-4ada-4482-fff1-4da629c677d5"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "running experiment for warp\n",
            "already clear\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.47354, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.47354\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.47354\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.47354\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.47354\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.47354\n",
            "\n",
            "Epoch 7: val_loss improved from 2.47354 to 2.18019, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.18019\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.18019\n",
            "\n",
            "Epoch 10: val_loss improved from 2.18019 to 2.14469, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.14469\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.14469\n",
            "\n",
            "Epoch 13: val_loss did not improve from 2.14469\n",
            "\n",
            "Epoch 14: val_loss did not improve from 2.14469\n",
            "\n",
            "Epoch 15: val_loss improved from 2.14469 to 2.08115, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 16: val_loss did not improve from 2.08115\n",
            "\n",
            "Epoch 17: val_loss improved from 2.08115 to 1.87824, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.87824\n",
            "\n",
            "Epoch 19: val_loss improved from 1.87824 to 1.74334, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.74334\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.74334\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.74334\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.74334\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.74334\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.74334\n",
            "\n",
            "Epoch 26: val_loss improved from 1.74334 to 1.54095, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 27: val_loss did not improve from 1.54095\n",
            "\n",
            "Epoch 28: val_loss did not improve from 1.54095\n",
            "\n",
            "Epoch 29: val_loss did not improve from 1.54095\n",
            "\n",
            "Epoch 30: val_loss did not improve from 1.54095\n",
            "\n",
            "Epoch 31: val_loss did not improve from 1.54095\n",
            "\n",
            "Epoch 32: val_loss did not improve from 1.54095\n",
            "\n",
            "Epoch 33: val_loss did not improve from 1.54095\n",
            "\n",
            "Epoch 34: val_loss did not improve from 1.54095\n",
            "6/6 [==============================] - 2s 149ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.42621, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.42621\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.42621\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.42621\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.42621\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.42621\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.42621\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.42621\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.42621\n",
            "6/6 [==============================] - 1s 55ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.31452, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.31452\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.31452\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.31452\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.31452\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.31452\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.31452\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.31452\n",
            "\n",
            "Epoch 9: val_loss improved from 2.31452 to 2.31255, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.56329, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.56329\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.56329\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.56329\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.56329\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.56329\n",
            "\n",
            "Epoch 7: val_loss improved from 2.56329 to 2.03725, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.03725\n",
            "\n",
            "Epoch 9: val_loss improved from 2.03725 to 1.95439, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.95439\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.95439\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.95439\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.95439\n",
            "\n",
            "Epoch 14: val_loss improved from 1.95439 to 1.52182, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.52182\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.52182\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.52182\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.52182\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.52182\n",
            "\n",
            "Epoch 20: val_loss improved from 1.52182 to 1.50142, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 21: val_loss improved from 1.50142 to 1.41853, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.41853\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.41853\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.41853\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.41853\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.41853\n",
            "\n",
            "Epoch 27: val_loss did not improve from 1.41853\n",
            "\n",
            "Epoch 28: val_loss did not improve from 1.41853\n",
            "\n",
            "Epoch 29: val_loss did not improve from 1.41853\n",
            "6/6 [==============================] - 2s 171ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.43488, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.43488\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.43488\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.43488\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.43488\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.43488\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.43488\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.43488\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.43488\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.47788, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.47788\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.47788\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.47788\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.47788\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.47788\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.47788\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.47788\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.47788\n",
            "6/6 [==============================] - 2s 152ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.42348, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.42348\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.42348\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.42348\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.42348\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.42348\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.42348\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.42348\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.42348\n",
            "6/6 [==============================] - 1s 57ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.50355, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.50355\n",
            "\n",
            "Epoch 3: val_loss improved from 2.50355 to 2.24475, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.24475\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.24475\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.24475\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.24475\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.24475\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.24475\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.24475\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.24475\n",
            "6/6 [==============================] - 1s 55ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.37701, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.37701\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.37701\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.37701\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.37701\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.37701\n",
            "\n",
            "Epoch 7: val_loss improved from 2.37701 to 2.35131, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss improved from 2.35131 to 2.27088, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.27088\n",
            "\n",
            "Epoch 10: val_loss improved from 2.27088 to 2.09289, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.09289\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.09289\n",
            "\n",
            "Epoch 13: val_loss did not improve from 2.09289\n",
            "\n",
            "Epoch 14: val_loss improved from 2.09289 to 1.73979, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.73979\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.73979\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.73979\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.73979\n",
            "\n",
            "Epoch 19: val_loss improved from 1.73979 to 1.56189, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.56189\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.56189\n",
            "\n",
            "Epoch 22: val_loss improved from 1.56189 to 1.50694, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 23: val_loss improved from 1.50694 to 1.36135, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.36135\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.36135\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.36135\n",
            "\n",
            "Epoch 27: val_loss did not improve from 1.36135\n",
            "\n",
            "Epoch 28: val_loss did not improve from 1.36135\n",
            "\n",
            "Epoch 29: val_loss did not improve from 1.36135\n",
            "\n",
            "Epoch 30: val_loss did not improve from 1.36135\n",
            "\n",
            "Epoch 31: val_loss did not improve from 1.36135\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.50006, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.50006\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.50006\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.50006\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.50006\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.50006\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.50006\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.50006\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.50006\n",
            "6/6 [==============================] - 1s 57ms/step\n",
            "done in 1848.83\n"
          ]
        }
      ],
      "source": [
        "run_experiments(names[4],backbone = backbone)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "background_save": true,
          "base_uri": "https://localhost:8080/"
        },
        "id": "5fmyMwaz1-dP",
        "outputId": "1bf8c336-8a71-4426-d4b6-7005689d12fe"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "running experiment for delay\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.61096, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.61096\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.61096\n",
            "\n",
            "Epoch 4: val_loss improved from 2.61096 to 2.42746, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss improved from 2.42746 to 2.30224, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.30224\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.30224\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.30224\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.30224\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.30224\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.30224\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.30224\n",
            "\n",
            "Epoch 13: val_loss did not improve from 2.30224\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.47672, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.47672\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.47672\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.47672\n",
            "\n",
            "Epoch 5: val_loss improved from 2.47672 to 2.39681, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.39681\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.39681\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.39681\n",
            "\n",
            "Epoch 9: val_loss improved from 2.39681 to 2.20446, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss improved from 2.20446 to 1.90520, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.90520\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.90520\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.90520\n",
            "\n",
            "Epoch 14: val_loss improved from 1.90520 to 1.88809, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 15: val_loss improved from 1.88809 to 1.77358, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 16: val_loss improved from 1.77358 to 1.51973, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.51973\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.51973\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.51973\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.51973\n",
            "\n",
            "Epoch 21: val_loss improved from 1.51973 to 1.51240, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.51240\n",
            "\n",
            "Epoch 23: val_loss improved from 1.51240 to 1.11670, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.11670\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.11670\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.11670\n",
            "\n",
            "Epoch 27: val_loss did not improve from 1.11670\n",
            "\n",
            "Epoch 28: val_loss did not improve from 1.11670\n",
            "\n",
            "Epoch 29: val_loss did not improve from 1.11670\n",
            "\n",
            "Epoch 30: val_loss did not improve from 1.11670\n",
            "\n",
            "Epoch 31: val_loss did not improve from 1.11670\n",
            "6/6 [==============================] - 1s 57ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.42144, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.42144\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.42144\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.42144\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.42144\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.42144\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.42144\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.42144\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.42144\n",
            "6/6 [==============================] - 1s 55ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.60081, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.60081\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.60081\n",
            "\n",
            "Epoch 4: val_loss improved from 2.60081 to 2.30890, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.30890\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.30890\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.30890\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.30890\n",
            "\n",
            "Epoch 9: val_loss improved from 2.30890 to 1.90318, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.90318\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.90318\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.90318\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.90318\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.90318\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.90318\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.90318\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.90318\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.67246, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.67246\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.67246\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.67246\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.67246\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.67246\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.67246\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.67246\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.67246\n",
            "6/6 [==============================] - 1s 55ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.48186, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.48186\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.48186\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.48186\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.48186\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.48186\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.48186\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.48186\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.48186\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.53697, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.53697\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.53697\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.53697\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.53697\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.53697\n",
            "\n",
            "Epoch 7: val_loss improved from 2.53697 to 2.09669, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.09669\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.09669\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.09669\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.09669\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.09669\n",
            "\n",
            "Epoch 13: val_loss did not improve from 2.09669\n",
            "\n",
            "Epoch 14: val_loss did not improve from 2.09669\n",
            "\n",
            "Epoch 15: val_loss improved from 2.09669 to 1.82874, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.82874\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.82874\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.82874\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.82874\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.82874\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.82874\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.82874\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.82874\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.42502, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.42502\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.42502\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.42502\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.42502\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.42502\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.42502\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.42502\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.42502\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.58416, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.58416\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.58416\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.58416\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.58416\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.58416\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.58416\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.58416\n",
            "\n",
            "Epoch 9: val_loss improved from 2.58416 to 2.29606, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.29606\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.29606\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.29606\n",
            "\n",
            "Epoch 13: val_loss did not improve from 2.29606\n",
            "\n",
            "Epoch 14: val_loss did not improve from 2.29606\n",
            "\n",
            "Epoch 15: val_loss did not improve from 2.29606\n",
            "\n",
            "Epoch 16: val_loss did not improve from 2.29606\n",
            "\n",
            "Epoch 17: val_loss did not improve from 2.29606\n",
            "6/6 [==============================] - 1s 55ms/step\n",
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.50416, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.50416\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.50416\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.50416\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.50416\n",
            "\n",
            "Epoch 6: val_loss improved from 2.50416 to 2.15214, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.15214\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.15214\n",
            "\n",
            "Epoch 9: val_loss improved from 2.15214 to 1.87504, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.87504\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.87504\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.87504\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.87504\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.87504\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.87504\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.87504\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.87504\n",
            "6/6 [==============================] - 1s 56ms/step\n",
            "done in 1812.43\n"
          ]
        }
      ],
      "source": [
        "run_experiments(names[5], backbone = backbone)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Xkcm4KzIvOfK"
      },
      "outputs": [],
      "source": [
        "run_experiments(names[6], backbone = backbone)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 201
        },
        "id": "91UN6FIsMnHO",
        "outputId": "a314129d-2b8c-46b7-d820-4da05b1fd851"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-9ad39e3a-414b-42ed-8631-d35da3226f13\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>slice_file_name</th>\n",
              "      <th>fold</th>\n",
              "      <th>class</th>\n",
              "      <th>source</th>\n",
              "      <th>dist</th>\n",
              "      <th>mixup</th>\n",
              "      <th>imixup</th>\n",
              "      <th>room</th>\n",
              "      <th>warp</th>\n",
              "      <th>delay</th>\n",
              "      <th>spectrum</th>\n",
              "      <th>all</th>\n",
              "      <th>image_only</th>\n",
              "      <th>audio_only</th>\n",
              "      <th>audio_only_nd</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>57320-0-0-39.wav</td>\n",
              "      <td>1</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>o</td>\n",
              "      <td>drilling</td>\n",
              "      <td>street_music</td>\n",
              "      <td>engine_idling</td>\n",
              "      <td>dog_bark</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>engine_idling</td>\n",
              "      <td>jackhammer</td>\n",
              "      <td>engine_idling</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>siren</td>\n",
              "      <td>street_music</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>134717-0-0-6.wav</td>\n",
              "      <td>1</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>o</td>\n",
              "      <td>drilling</td>\n",
              "      <td>street_music</td>\n",
              "      <td>engine_idling</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>jackhammer</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>jackhammer</td>\n",
              "      <td>engine_idling</td>\n",
              "      <td>siren</td>\n",
              "      <td>street_music</td>\n",
              "      <td>street_music</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>57320-0-0-31.wav</td>\n",
              "      <td>1</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>o</td>\n",
              "      <td>drilling</td>\n",
              "      <td>street_music</td>\n",
              "      <td>children_playing</td>\n",
              "      <td>dog_bark</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>engine_idling</td>\n",
              "      <td>jackhammer</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>dog_bark</td>\n",
              "      <td>siren</td>\n",
              "      <td>street_music</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>134717-0-0-15.wav</td>\n",
              "      <td>1</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>o</td>\n",
              "      <td>drilling</td>\n",
              "      <td>street_music</td>\n",
              "      <td>engine_idling</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>engine_idling</td>\n",
              "      <td>jackhammer</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>siren</td>\n",
              "      <td>street_music</td>\n",
              "      <td>street_music</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>177621-0-0-27.wav</td>\n",
              "      <td>1</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>o</td>\n",
              "      <td>drilling</td>\n",
              "      <td>street_music</td>\n",
              "      <td>engine_idling</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>air_conditioner</td>\n",
              "      <td>engine_idling</td>\n",
              "      <td>jackhammer</td>\n",
              "      <td>engine_idling</td>\n",
              "      <td>engine_idling</td>\n",
              "      <td>street_music</td>\n",
              "      <td>street_music</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-9ad39e3a-414b-42ed-8631-d35da3226f13')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-9ad39e3a-414b-42ed-8631-d35da3226f13 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-9ad39e3a-414b-42ed-8631-d35da3226f13');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-0d349de4-85b7-4c65-b0f4-bf22c5f43447\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-0d349de4-85b7-4c65-b0f4-bf22c5f43447')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-0d349de4-85b7-4c65-b0f4-bf22c5f43447 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "text/plain": [
              "     slice_file_name fold            class source      dist         mixup  \\\n",
              "0   57320-0-0-39.wav    1  air_conditioner      o  drilling  street_music   \n",
              "1   134717-0-0-6.wav    1  air_conditioner      o  drilling  street_music   \n",
              "2   57320-0-0-31.wav    1  air_conditioner      o  drilling  street_music   \n",
              "3  134717-0-0-15.wav    1  air_conditioner      o  drilling  street_music   \n",
              "4  177621-0-0-27.wav    1  air_conditioner      o  drilling  street_music   \n",
              "\n",
              "             imixup             room             warp            delay  \\\n",
              "0     engine_idling         dog_bark  air_conditioner    engine_idling   \n",
              "1     engine_idling  air_conditioner       jackhammer  air_conditioner   \n",
              "2  children_playing         dog_bark  air_conditioner    engine_idling   \n",
              "3     engine_idling  air_conditioner  air_conditioner    engine_idling   \n",
              "4     engine_idling  air_conditioner  air_conditioner    engine_idling   \n",
              "\n",
              "     spectrum              all       image_only    audio_only audio_only_nd  \n",
              "0  jackhammer    engine_idling  air_conditioner         siren  street_music  \n",
              "1  jackhammer    engine_idling            siren  street_music  street_music  \n",
              "2  jackhammer  air_conditioner         dog_bark         siren  street_music  \n",
              "3  jackhammer  air_conditioner            siren  street_music  street_music  \n",
              "4  jackhammer    engine_idling    engine_idling  street_music  street_music  "
            ]
          },
          "execution_count": 19,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "pd.read_pickle(path + 'results.pkl').head(5)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "F7vwCdpRHK3B"
      },
      "outputs": [],
      "source": [
        "#Now it is time to see how various combinations of augmentation techniques are performing.\n",
        "#The following cell is creating a function similar to run_experiments but for group of augmentation techniques\n",
        "\n",
        "def run_grouped_experiments(group, group_name, backbone = backbone, batch_size = 64):\n",
        "  \"\"\" This function is running experiments for a group of augmentation techniques, very similar to run_experiments function.\n",
        "  group is a list of augmentation techniques to be used.\n",
        "  group_name is a name of the group, which will be used to name the column in the results dataframe.\n",
        "  backbone is a type of backbone to be used to build a model.\n",
        "  batch_size is a batch size to be used for training.\n",
        "  \"\"\"\n",
        "  start = time.time()\n",
        "  results = pd.read_pickle(path + 'results.pkl')\n",
        "  dfs = []\n",
        "  for n in group:\n",
        "      if n != 'spectrum':\n",
        "          dfs.append(build_df(n)[1])\n",
        "      else:\n",
        "          dfs.append(build_df(n)[0])\n",
        "  df = pd.concat(dfs, ignore_index=True, axis=0)\n",
        "  r = pd.DataFrame(columns=['slice_file_name', 'fold', 'source', name])\n",
        "  for fold in results.fold.unique():\n",
        "      train = df[df['fold'] != fold]\n",
        "      val_fold = train.fold.sample(n=1).values[0]\n",
        "      val = train[train['fold'] == val_fold]\n",
        "      val = val[val.source == 'o']\n",
        "      train = train[train['fold'] != val_fold]\n",
        "      test = df[df['fold'] == fold]\n",
        "      test = test[test.source == 'o']\n",
        "      train_x = np.stack([x for x in train['r'].values])\n",
        "      train_y = train[classes]\n",
        "      test = df[df['fold'] == fold]\n",
        "      test_x = np.stack([x for x in test['r'].values])\n",
        "      test_y = test[classes]\n",
        "      val_x = np.stack([x for x in val['r'].values])\n",
        "      val_y = val[classes]\n",
        "      test_link = test[['slice_file_name', 'fold', 'source']]\n",
        "      pred = run_model(train_x, train_y, test_x, test_y, val_x, val_y, backbone = backbone)\n",
        "      test_link[name] = pd.DataFrame(\n",
        "          pred, columns=classes).idxmax(axis=1).to_numpy()\n",
        "      r = pd.concat(\n",
        "          [r, test_link[['slice_file_name', 'fold', 'source', name]]], ignore_index=True)\n",
        "  results = pd.merge(results, r,  how='left', left_on=[\n",
        "                    'slice_file_name', 'fold', 'source'], right_on=['slice_file_name', 'fold', 'source'])\n",
        "  results.to_pickle(path + \"results.pkl\")\n",
        "  end = time.time()\n",
        "  print(f'done in {str(round(end - start,2))}')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "background_save": true,
          "base_uri": "https://localhost:8080/"
        },
        "id": "W7vsEds1hLSS",
        "outputId": "85399f81-3c6d-43bf-ca51-4b905243654b"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.60920, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.60920\n",
            "\n",
            "Epoch 3: val_loss improved from 2.60920 to 2.27392, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss improved from 2.27392 to 2.26172, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.26172\n",
            "\n",
            "Epoch 6: val_loss improved from 2.26172 to 1.88169, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss improved from 1.88169 to 1.70723, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss improved from 1.70723 to 1.42418, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.42418\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.42418\n",
            "\n",
            "Epoch 11: val_loss improved from 1.42418 to 0.98346, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 12: val_loss did not improve from 0.98346\n",
            "\n",
            "Epoch 13: val_loss did not improve from 0.98346\n",
            "\n",
            "Epoch 14: val_loss did not improve from 0.98346\n",
            "\n",
            "Epoch 15: val_loss did not improve from 0.98346\n",
            "\n",
            "Epoch 16: val_loss did not improve from 0.98346\n",
            "\n",
            "Epoch 17: val_loss did not improve from 0.98346\n",
            "\n",
            "Epoch 18: val_loss did not improve from 0.98346\n",
            "\n",
            "Epoch 19: val_loss did not improve from 0.98346\n",
            "20/20 [==============================] - 3s 54ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.53897, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.53897\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.53897\n",
            "\n",
            "Epoch 4: val_loss improved from 2.53897 to 1.93504, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 1.93504\n",
            "\n",
            "Epoch 6: val_loss improved from 1.93504 to 1.92959, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss improved from 1.92959 to 1.15629, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.15629\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.15629\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.15629\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.15629\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.15629\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.15629\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.15629\n",
            "\n",
            "Epoch 15: val_loss improved from 1.15629 to 0.97540, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 16: val_loss did not improve from 0.97540\n",
            "\n",
            "Epoch 17: val_loss did not improve from 0.97540\n",
            "\n",
            "Epoch 18: val_loss improved from 0.97540 to 0.94716, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 19: val_loss did not improve from 0.94716\n",
            "\n",
            "Epoch 20: val_loss did not improve from 0.94716\n",
            "\n",
            "Epoch 21: val_loss did not improve from 0.94716\n",
            "\n",
            "Epoch 22: val_loss did not improve from 0.94716\n",
            "\n",
            "Epoch 23: val_loss did not improve from 0.94716\n",
            "\n",
            "Epoch 24: val_loss did not improve from 0.94716\n",
            "\n",
            "Epoch 25: val_loss did not improve from 0.94716\n",
            "\n",
            "Epoch 26: val_loss improved from 0.94716 to 0.90020, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 27: val_loss did not improve from 0.90020\n",
            "\n",
            "Epoch 28: val_loss did not improve from 0.90020\n",
            "\n",
            "Epoch 29: val_loss did not improve from 0.90020\n",
            "\n",
            "Epoch 30: val_loss did not improve from 0.90020\n",
            "\n",
            "Epoch 31: val_loss did not improve from 0.90020\n",
            "\n",
            "Epoch 32: val_loss did not improve from 0.90020\n",
            "\n",
            "Epoch 33: val_loss did not improve from 0.90020\n",
            "\n",
            "Epoch 34: val_loss did not improve from 0.90020\n",
            "20/20 [==============================] - 2s 58ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.89558, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 2.89558 to 2.22861, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.22861\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.22861\n",
            "\n",
            "Epoch 5: val_loss improved from 2.22861 to 1.59872, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 1.59872\n",
            "\n",
            "Epoch 7: val_loss improved from 1.59872 to 1.38568, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.38568\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.38568\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.38568\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.38568\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.38568\n",
            "\n",
            "Epoch 13: val_loss improved from 1.38568 to 0.90127, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 14: val_loss did not improve from 0.90127\n",
            "\n",
            "Epoch 15: val_loss did not improve from 0.90127\n",
            "\n",
            "Epoch 16: val_loss did not improve from 0.90127\n",
            "\n",
            "Epoch 17: val_loss did not improve from 0.90127\n",
            "\n",
            "Epoch 18: val_loss did not improve from 0.90127\n",
            "\n",
            "Epoch 19: val_loss did not improve from 0.90127\n",
            "\n",
            "Epoch 20: val_loss did not improve from 0.90127\n",
            "\n",
            "Epoch 21: val_loss did not improve from 0.90127\n",
            "20/20 [==============================] - 2s 35ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.82183, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 2.82183 to 2.36483, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss improved from 2.36483 to 1.78642, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss did not improve from 1.78642\n",
            "\n",
            "Epoch 5: val_loss did not improve from 1.78642\n",
            "\n",
            "Epoch 6: val_loss did not improve from 1.78642\n",
            "\n",
            "Epoch 7: val_loss improved from 1.78642 to 1.68963, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss improved from 1.68963 to 1.49018, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.49018\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.49018\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.49018\n",
            "\n",
            "Epoch 12: val_loss improved from 1.49018 to 1.22413, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.22413\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.22413\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.22413\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.22413\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.22413\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.22413\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.22413\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.22413\n",
            "21/21 [==============================] - 2s 50ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 3.08362, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 3.08362 to 2.91068, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss improved from 2.91068 to 2.79204, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.79204\n",
            "\n",
            "Epoch 5: val_loss improved from 2.79204 to 2.59458, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss improved from 2.59458 to 1.60492, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss did not improve from 1.60492\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.60492\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.60492\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.60492\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.60492\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.60492\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.60492\n",
            "\n",
            "Epoch 14: val_loss improved from 1.60492 to 1.56570, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.56570\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.56570\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.56570\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.56570\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.56570\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.56570\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.56570\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.56570\n",
            "21/21 [==============================] - 2s 35ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.96916, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 2.96916 to 2.22873, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss improved from 2.22873 to 1.82231, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss did not improve from 1.82231\n",
            "\n",
            "Epoch 5: val_loss improved from 1.82231 to 1.79814, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 1.79814\n",
            "\n",
            "Epoch 7: val_loss did not improve from 1.79814\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.79814\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.79814\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.79814\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.79814\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.79814\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.79814\n",
            "20/20 [==============================] - 2s 55ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.58550, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 2.58550 to 2.05623, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.05623\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.05623\n",
            "\n",
            "Epoch 5: val_loss improved from 2.05623 to 1.57424, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 1.57424\n",
            "\n",
            "Epoch 7: val_loss did not improve from 1.57424\n",
            "\n",
            "Epoch 8: val_loss improved from 1.57424 to 1.18104, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.18104\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.18104\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.18104\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.18104\n",
            "\n",
            "Epoch 13: val_loss improved from 1.18104 to 1.12644, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.12644\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.12644\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.12644\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.12644\n",
            "\n",
            "Epoch 18: val_loss improved from 1.12644 to 1.11888, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.11888\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.11888\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.11888\n",
            "21/21 [==============================] - 2s 41ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.85051, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 2.85051 to 2.56115, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss improved from 2.56115 to 1.91708, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss improved from 1.91708 to 1.85258, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 1.85258\n",
            "\n",
            "Epoch 6: val_loss did not improve from 1.85258\n",
            "\n",
            "Epoch 7: val_loss improved from 1.85258 to 1.43910, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.43910\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.43910\n",
            "\n",
            "Epoch 10: val_loss improved from 1.43910 to 1.43485, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.43485\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.43485\n",
            "\n",
            "Epoch 13: val_loss improved from 1.43485 to 1.16783, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.16783\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.16783\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.16783\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.16783\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.16783\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.16783\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.16783\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.16783\n",
            "20/20 [==============================] - 2s 48ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.64238, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.64238\n",
            "\n",
            "Epoch 3: val_loss improved from 2.64238 to 1.97643, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss did not improve from 1.97643\n",
            "\n",
            "Epoch 5: val_loss did not improve from 1.97643\n",
            "\n",
            "Epoch 6: val_loss improved from 1.97643 to 1.75175, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss did not improve from 1.75175\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.75175\n",
            "\n",
            "Epoch 9: val_loss improved from 1.75175 to 1.37991, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.37991\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.37991\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.37991\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.37991\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.37991\n",
            "\n",
            "Epoch 15: val_loss improved from 1.37991 to 1.22912, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.22912\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.22912\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.22912\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.22912\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.22912\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.22912\n",
            "\n",
            "Epoch 22: val_loss improved from 1.22912 to 1.22664, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.22664\n",
            "20/20 [==============================] - 2s 34ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.95069, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 2.95069 to 2.77147, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss improved from 2.77147 to 2.36732, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss improved from 2.36732 to 2.28434, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.28434\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.28434\n",
            "\n",
            "Epoch 7: val_loss improved from 2.28434 to 1.51696, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.51696\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.51696\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.51696\n",
            "\n",
            "Epoch 11: val_loss improved from 1.51696 to 1.38651, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.38651\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.38651\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.38651\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.38651\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.38651\n",
            "\n",
            "Epoch 17: val_loss improved from 1.38651 to 1.37893, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.37893\n",
            "\n",
            "Epoch 19: val_loss improved from 1.37893 to 1.29313, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.29313\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.29313\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.29313\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.29313\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.29313\n",
            "\n",
            "Epoch 25: val_loss improved from 1.29313 to 0.86720, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 26: val_loss did not improve from 0.86720\n",
            "\n",
            "Epoch 27: val_loss did not improve from 0.86720\n",
            "\n",
            "Epoch 28: val_loss did not improve from 0.86720\n",
            "\n",
            "Epoch 29: val_loss did not improve from 0.86720\n",
            "\n",
            "Epoch 30: val_loss did not improve from 0.86720\n",
            "\n",
            "Epoch 31: val_loss did not improve from 0.86720\n",
            "\n",
            "Epoch 32: val_loss did not improve from 0.86720\n",
            "\n",
            "Epoch 33: val_loss did not improve from 0.86720\n",
            "20/20 [==============================] - 2s 35ms/step\n",
            "done in 3691.47\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        }
      ],
      "source": [
        "run_grouped_experiments(names,\"all\",backbone = backbone, batch_size = 64)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "background_save": true
        },
        "id": "PLEwXhSvhbe_",
        "outputId": "b13dd899-d123-47f9-fe2d-01e1c2b8d719"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.51082, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.51082\n",
            "\n",
            "Epoch 3: val_loss improved from 2.51082 to 2.28577, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss improved from 2.28577 to 2.02350, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.02350\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.02350\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.02350\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.02350\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.02350\n",
            "\n",
            "Epoch 10: val_loss improved from 2.02350 to 1.70840, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.70840\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.70840\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.70840\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.70840\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.70840\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.70840\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.70840\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.70840\n",
            "9/9 [==============================] - 1s 67ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.68871, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.68871\n",
            "\n",
            "Epoch 3: val_loss improved from 2.68871 to 2.66939, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss improved from 2.66939 to 2.45113, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss improved from 2.45113 to 2.17658, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.17658\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.17658\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.17658\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.17658\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.17658\n",
            "\n",
            "Epoch 11: val_loss improved from 2.17658 to 1.67929, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.67929\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.67929\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.67929\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.67929\n",
            "\n",
            "Epoch 16: val_loss improved from 1.67929 to 1.13372, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.13372\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.13372\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.13372\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.13372\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.13372\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.13372\n",
            "\n",
            "Epoch 23: val_loss improved from 1.13372 to 1.09509, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.09509\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.09509\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.09509\n",
            "\n",
            "Epoch 27: val_loss did not improve from 1.09509\n",
            "\n",
            "Epoch 28: val_loss improved from 1.09509 to 0.85142, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 29: val_loss did not improve from 0.85142\n",
            "\n",
            "Epoch 30: val_loss did not improve from 0.85142\n",
            "\n",
            "Epoch 31: val_loss did not improve from 0.85142\n",
            "\n",
            "Epoch 32: val_loss did not improve from 0.85142\n",
            "\n",
            "Epoch 33: val_loss did not improve from 0.85142\n",
            "\n",
            "Epoch 34: val_loss did not improve from 0.85142\n",
            "\n",
            "Epoch 35: val_loss did not improve from 0.85142\n",
            "\n",
            "Epoch 36: val_loss did not improve from 0.85142\n",
            "9/9 [==============================] - 1s 72ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.92779, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.92779\n",
            "\n",
            "Epoch 3: val_loss improved from 2.92779 to 2.27046, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.27046\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.27046\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.27046\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.27046\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.27046\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.27046\n",
            "\n",
            "Epoch 10: val_loss improved from 2.27046 to 2.15946, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.15946\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.15946\n",
            "\n",
            "Epoch 13: val_loss did not improve from 2.15946\n",
            "\n",
            "Epoch 14: val_loss did not improve from 2.15946\n",
            "\n",
            "Epoch 15: val_loss did not improve from 2.15946\n",
            "\n",
            "Epoch 16: val_loss did not improve from 2.15946\n",
            "\n",
            "Epoch 17: val_loss improved from 2.15946 to 1.78222, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 18: val_loss improved from 1.78222 to 1.65412, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 19: val_loss improved from 1.65412 to 1.19479, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.19479\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.19479\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.19479\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.19479\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.19479\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.19479\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.19479\n",
            "\n",
            "Epoch 27: val_loss did not improve from 1.19479\n",
            "9/9 [==============================] - 1s 36ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.66959, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.66959\n",
            "\n",
            "Epoch 3: val_loss improved from 2.66959 to 2.51437, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss improved from 2.51437 to 2.27950, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss improved from 2.27950 to 2.06937, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.06937\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.06937\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.06937\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.06937\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.06937\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.06937\n",
            "\n",
            "Epoch 12: val_loss improved from 2.06937 to 1.49287, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.49287\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.49287\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.49287\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.49287\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.49287\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.49287\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.49287\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.49287\n",
            "9/9 [==============================] - 1s 37ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.66562, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.66562\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.66562\n",
            "\n",
            "Epoch 4: val_loss improved from 2.66562 to 2.13075, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.13075\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.13075\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.13075\n",
            "\n",
            "Epoch 8: val_loss improved from 2.13075 to 1.85476, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.85476\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.85476\n",
            "\n",
            "Epoch 11: val_loss improved from 1.85476 to 1.54780, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.54780\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.54780\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.54780\n",
            "\n",
            "Epoch 15: val_loss improved from 1.54780 to 1.44871, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 16: val_loss improved from 1.44871 to 1.37473, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.37473\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.37473\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.37473\n",
            "\n",
            "Epoch 20: val_loss improved from 1.37473 to 0.93695, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 21: val_loss did not improve from 0.93695\n",
            "\n",
            "Epoch 22: val_loss did not improve from 0.93695\n",
            "\n",
            "Epoch 23: val_loss did not improve from 0.93695\n",
            "\n",
            "Epoch 24: val_loss did not improve from 0.93695\n",
            "\n",
            "Epoch 25: val_loss did not improve from 0.93695\n",
            "\n",
            "Epoch 26: val_loss did not improve from 0.93695\n",
            "\n",
            "Epoch 27: val_loss did not improve from 0.93695\n",
            "\n",
            "Epoch 28: val_loss did not improve from 0.93695\n",
            "9/9 [==============================] - 1s 38ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.66421, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.66421\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.66421\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.66421\n",
            "\n",
            "Epoch 5: val_loss improved from 2.66421 to 2.56673, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss improved from 2.56673 to 2.18362, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss improved from 2.18362 to 1.85091, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss improved from 1.85091 to 1.82681, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.82681\n",
            "\n",
            "Epoch 10: val_loss improved from 1.82681 to 1.62001, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.62001\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.62001\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.62001\n",
            "\n",
            "Epoch 14: val_loss improved from 1.62001 to 1.58560, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.58560\n",
            "\n",
            "Epoch 16: val_loss improved from 1.58560 to 1.46112, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.46112\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.46112\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.46112\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.46112\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.46112\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.46112\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.46112\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.46112\n",
            "9/9 [==============================] - 1s 70ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.30652, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.30652\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.30652\n",
            "\n",
            "Epoch 4: val_loss improved from 2.30652 to 2.23937, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.23937\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.23937\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.23937\n",
            "\n",
            "Epoch 8: val_loss improved from 2.23937 to 2.13742, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss improved from 2.13742 to 1.99169, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.99169\n",
            "\n",
            "Epoch 11: val_loss improved from 1.99169 to 1.95917, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.95917\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.95917\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.95917\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.95917\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.95917\n",
            "\n",
            "Epoch 17: val_loss improved from 1.95917 to 1.66949, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.66949\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.66949\n",
            "\n",
            "Epoch 20: val_loss improved from 1.66949 to 1.52914, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.52914\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.52914\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.52914\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.52914\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.52914\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.52914\n",
            "\n",
            "Epoch 27: val_loss did not improve from 1.52914\n",
            "\n",
            "Epoch 28: val_loss improved from 1.52914 to 1.49259, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 29: val_loss did not improve from 1.49259\n",
            "\n",
            "Epoch 30: val_loss did not improve from 1.49259\n",
            "\n",
            "Epoch 31: val_loss did not improve from 1.49259\n",
            "\n",
            "Epoch 32: val_loss did not improve from 1.49259\n",
            "\n",
            "Epoch 33: val_loss did not improve from 1.49259\n",
            "\n",
            "Epoch 34: val_loss did not improve from 1.49259\n",
            "\n",
            "Epoch 35: val_loss did not improve from 1.49259\n",
            "\n",
            "Epoch 36: val_loss did not improve from 1.49259\n",
            "9/9 [==============================] - 1s 74ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.64719, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.64719\n",
            "\n",
            "Epoch 3: val_loss improved from 2.64719 to 2.16150, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.16150\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.16150\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.16150\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.16150\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.16150\n",
            "\n",
            "Epoch 9: val_loss improved from 2.16150 to 1.58694, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.58694\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.58694\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.58694\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.58694\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.58694\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.58694\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.58694\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.58694\n",
            "9/9 [==============================] - 1s 60ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.64763, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.64763\n",
            "\n",
            "Epoch 3: val_loss improved from 2.64763 to 2.32288, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.32288\n",
            "\n",
            "Epoch 5: val_loss improved from 2.32288 to 2.14427, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss improved from 2.14427 to 1.78796, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss did not improve from 1.78796\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.78796\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.78796\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.78796\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.78796\n",
            "\n",
            "Epoch 12: val_loss improved from 1.78796 to 1.70517, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.70517\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.70517\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.70517\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.70517\n",
            "\n",
            "Epoch 17: val_loss improved from 1.70517 to 1.45323, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.45323\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.45323\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.45323\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.45323\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.45323\n",
            "\n",
            "Epoch 23: val_loss improved from 1.45323 to 1.38887, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.38887\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.38887\n",
            "\n",
            "Epoch 26: val_loss improved from 1.38887 to 1.31013, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 27: val_loss did not improve from 1.31013\n",
            "\n",
            "Epoch 28: val_loss did not improve from 1.31013\n",
            "\n",
            "Epoch 29: val_loss did not improve from 1.31013\n",
            "\n",
            "Epoch 30: val_loss improved from 1.31013 to 1.24721, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 31: val_loss did not improve from 1.24721\n",
            "\n",
            "Epoch 32: val_loss did not improve from 1.24721\n",
            "\n",
            "Epoch 33: val_loss did not improve from 1.24721\n",
            "\n",
            "Epoch 34: val_loss did not improve from 1.24721\n",
            "\n",
            "Epoch 35: val_loss did not improve from 1.24721\n",
            "\n",
            "Epoch 36: val_loss did not improve from 1.24721\n",
            "\n",
            "Epoch 37: val_loss improved from 1.24721 to 1.15894, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 38: val_loss did not improve from 1.15894\n",
            "\n",
            "Epoch 39: val_loss did not improve from 1.15894\n",
            "\n",
            "Epoch 40: val_loss did not improve from 1.15894\n",
            "\n",
            "Epoch 41: val_loss did not improve from 1.15894\n",
            "\n",
            "Epoch 42: val_loss did not improve from 1.15894\n",
            "\n",
            "Epoch 43: val_loss did not improve from 1.15894\n",
            "\n",
            "Epoch 44: val_loss did not improve from 1.15894\n",
            "\n",
            "Epoch 45: val_loss did not improve from 1.15894\n",
            "9/9 [==============================] - 1s 37ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.56303, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.56303\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.56303\n",
            "\n",
            "Epoch 4: val_loss improved from 2.56303 to 2.44728, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss improved from 2.44728 to 2.36759, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss improved from 2.36759 to 1.76816, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss did not improve from 1.76816\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.76816\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.76816\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.76816\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.76816\n",
            "\n",
            "Epoch 12: val_loss improved from 1.76816 to 1.12982, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.12982\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.12982\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.12982\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.12982\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.12982\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.12982\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.12982\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.12982\n",
            "9/9 [==============================] - 1s 37ms/step\n",
            "done in 2207.39\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        }
      ],
      "source": [
        "run_grouped_experiments(['imixup','warp','spectrum'],\"image_only\",backbone = backbone, batch_size = 64)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yv2Y330LhhZS",
        "outputId": "2a192a71-f84b-46f7-8e0d-90442f1977fd"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.72122, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 2.72122 to 2.46126, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.46126\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.46126\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.46126\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.46126\n",
            "\n",
            "Epoch 7: val_loss improved from 2.46126 to 1.32940, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.32940\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.32940\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.32940\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.32940\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.32940\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.32940\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.32940\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.32940\n",
            "15/15 [==============================] - 2s 43ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.63730, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.63730\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.63730\n",
            "\n",
            "Epoch 4: val_loss improved from 2.63730 to 2.39999, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.39999\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.39999\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.39999\n",
            "\n",
            "Epoch 8: val_loss improved from 2.39999 to 1.76946, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss improved from 1.76946 to 1.50228, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.50228\n",
            "\n",
            "Epoch 11: val_loss improved from 1.50228 to 1.33114, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 12: val_loss improved from 1.33114 to 1.12491, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.12491\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.12491\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.12491\n",
            "\n",
            "Epoch 16: val_loss improved from 1.12491 to 0.99971, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 17: val_loss did not improve from 0.99971\n",
            "\n",
            "Epoch 18: val_loss did not improve from 0.99971\n",
            "\n",
            "Epoch 19: val_loss did not improve from 0.99971\n",
            "\n",
            "Epoch 20: val_loss did not improve from 0.99971\n",
            "\n",
            "Epoch 21: val_loss did not improve from 0.99971\n",
            "\n",
            "Epoch 22: val_loss did not improve from 0.99971\n",
            "\n",
            "Epoch 23: val_loss did not improve from 0.99971\n",
            "\n",
            "Epoch 24: val_loss did not improve from 0.99971\n",
            "15/15 [==============================] - 2s 48ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.70159, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 2.70159 to 2.29252, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.29252\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.29252\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.29252\n",
            "\n",
            "Epoch 6: val_loss improved from 2.29252 to 2.01853, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.01853\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.01853\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.01853\n",
            "\n",
            "Epoch 10: val_loss improved from 2.01853 to 1.49401, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.49401\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.49401\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.49401\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.49401\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.49401\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.49401\n",
            "\n",
            "Epoch 17: val_loss improved from 1.49401 to 1.48200, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.48200\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.48200\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.48200\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.48200\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.48200\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.48200\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.48200\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.48200\n",
            "15/15 [==============================] - 1s 35ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.64496, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 2.64496 to 2.54594, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss improved from 2.54594 to 2.18218, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss improved from 2.18218 to 1.97617, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 1.97617\n",
            "\n",
            "Epoch 6: val_loss did not improve from 1.97617\n",
            "\n",
            "Epoch 7: val_loss improved from 1.97617 to 1.37319, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.37319\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.37319\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.37319\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.37319\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.37319\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.37319\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.37319\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.37319\n",
            "15/15 [==============================] - 1s 36ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.69550, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 2.69550 to 2.43578, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.43578\n",
            "\n",
            "Epoch 4: val_loss improved from 2.43578 to 2.42422, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss improved from 2.42422 to 1.91433, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 1.91433\n",
            "\n",
            "Epoch 7: val_loss did not improve from 1.91433\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.91433\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.91433\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.91433\n",
            "\n",
            "Epoch 11: val_loss improved from 1.91433 to 1.85903, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.85903\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.85903\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.85903\n",
            "\n",
            "Epoch 15: val_loss improved from 1.85903 to 1.61895, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 16: val_loss improved from 1.61895 to 1.61443, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.61443\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.61443\n",
            "\n",
            "Epoch 19: val_loss improved from 1.61443 to 1.44820, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.44820\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.44820\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.44820\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.44820\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.44820\n",
            "\n",
            "Epoch 25: val_loss improved from 1.44820 to 1.41420, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.41420\n",
            "\n",
            "Epoch 27: val_loss did not improve from 1.41420\n",
            "\n",
            "Epoch 28: val_loss did not improve from 1.41420\n",
            "\n",
            "Epoch 29: val_loss did not improve from 1.41420\n",
            "\n",
            "Epoch 30: val_loss did not improve from 1.41420\n",
            "\n",
            "Epoch 31: val_loss did not improve from 1.41420\n",
            "\n",
            "Epoch 32: val_loss did not improve from 1.41420\n",
            "\n",
            "Epoch 33: val_loss did not improve from 1.41420\n",
            "15/15 [==============================] - 1s 37ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.80172, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 2.80172 to 2.37203, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss improved from 2.37203 to 1.85049, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss did not improve from 1.85049\n",
            "\n",
            "Epoch 5: val_loss improved from 1.85049 to 1.63543, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 1.63543\n",
            "\n",
            "Epoch 7: val_loss improved from 1.63543 to 0.99580, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 0.99580\n",
            "\n",
            "Epoch 9: val_loss did not improve from 0.99580\n",
            "\n",
            "Epoch 10: val_loss did not improve from 0.99580\n",
            "\n",
            "Epoch 11: val_loss did not improve from 0.99580\n",
            "\n",
            "Epoch 12: val_loss did not improve from 0.99580\n",
            "\n",
            "Epoch 13: val_loss did not improve from 0.99580\n",
            "\n",
            "Epoch 14: val_loss did not improve from 0.99580\n",
            "\n",
            "Epoch 15: val_loss did not improve from 0.99580\n",
            "15/15 [==============================] - 2s 46ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.92730, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 2.92730 to 2.65838, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.65838\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.65838\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.65838\n",
            "\n",
            "Epoch 6: val_loss improved from 2.65838 to 2.01519, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss improved from 2.01519 to 1.86426, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.86426\n",
            "\n",
            "Epoch 9: val_loss improved from 1.86426 to 1.71108, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.71108\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.71108\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.71108\n",
            "\n",
            "Epoch 13: val_loss improved from 1.71108 to 1.46611, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.46611\n",
            "\n",
            "Epoch 15: val_loss improved from 1.46611 to 1.41709, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.41709\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.41709\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.41709\n",
            "\n",
            "Epoch 19: val_loss improved from 1.41709 to 1.23129, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.23129\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.23129\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.23129\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.23129\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.23129\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.23129\n",
            "\n",
            "Epoch 26: val_loss improved from 1.23129 to 1.09256, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 27: val_loss did not improve from 1.09256\n",
            "\n",
            "Epoch 28: val_loss did not improve from 1.09256\n",
            "\n",
            "Epoch 29: val_loss did not improve from 1.09256\n",
            "\n",
            "Epoch 30: val_loss did not improve from 1.09256\n",
            "\n",
            "Epoch 31: val_loss did not improve from 1.09256\n",
            "\n",
            "Epoch 32: val_loss did not improve from 1.09256\n",
            "\n",
            "Epoch 33: val_loss did not improve from 1.09256\n",
            "\n",
            "Epoch 34: val_loss improved from 1.09256 to 1.05058, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 35: val_loss did not improve from 1.05058\n",
            "\n",
            "Epoch 36: val_loss did not improve from 1.05058\n",
            "\n",
            "Epoch 37: val_loss did not improve from 1.05058\n",
            "\n",
            "Epoch 38: val_loss did not improve from 1.05058\n",
            "\n",
            "Epoch 39: val_loss did not improve from 1.05058\n",
            "\n",
            "Epoch 40: val_loss did not improve from 1.05058\n",
            "\n",
            "Epoch 41: val_loss did not improve from 1.05058\n",
            "\n",
            "Epoch 42: val_loss did not improve from 1.05058\n",
            "15/15 [==============================] - 2s 50ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.78496, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 2.78496 to 2.70763, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.70763\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.70763\n",
            "\n",
            "Epoch 5: val_loss improved from 2.70763 to 2.30237, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.30237\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.30237\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.30237\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.30237\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.30237\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.30237\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.30237\n",
            "\n",
            "Epoch 13: val_loss improved from 2.30237 to 2.30009, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "14/14 [==============================] - 1s 37ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.73104, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 2.73104 to 2.25768, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.25768\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.25768\n",
            "\n",
            "Epoch 5: val_loss improved from 2.25768 to 1.84531, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 1.84531\n",
            "\n",
            "Epoch 7: val_loss improved from 1.84531 to 1.75718, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss improved from 1.75718 to 1.66718, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 9: val_loss improved from 1.66718 to 1.55075, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss improved from 1.55075 to 1.52324, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.52324\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.52324\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.52324\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.52324\n",
            "\n",
            "Epoch 15: val_loss improved from 1.52324 to 1.31657, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.31657\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.31657\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.31657\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.31657\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.31657\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.31657\n",
            "\n",
            "Epoch 22: val_loss improved from 1.31657 to 1.14140, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.14140\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.14140\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.14140\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.14140\n",
            "\n",
            "Epoch 27: val_loss did not improve from 1.14140\n",
            "\n",
            "Epoch 28: val_loss did not improve from 1.14140\n",
            "\n",
            "Epoch 29: val_loss did not improve from 1.14140\n",
            "\n",
            "Epoch 30: val_loss did not improve from 1.14140\n",
            "14/14 [==============================] - 1s 35ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.68334, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.68334\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.68334\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.68334\n",
            "\n",
            "Epoch 5: val_loss improved from 2.68334 to 2.53535, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss improved from 2.53535 to 2.25556, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.25556\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.25556\n",
            "\n",
            "Epoch 9: val_loss improved from 2.25556 to 1.98942, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss improved from 1.98942 to 1.97670, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.97670\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.97670\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.97670\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.97670\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.97670\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.97670\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.97670\n",
            "\n",
            "Epoch 18: val_loss improved from 1.97670 to 1.64970, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.64970\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.64970\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.64970\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.64970\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.64970\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.64970\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.64970\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.64970\n",
            "14/14 [==============================] - 1s 35ms/step\n",
            "done in 2865.75\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        }
      ],
      "source": [
        "run_grouped_experiments(['dist', 'mixup', 'room','delay', 'spectrum'],\"audio_only\",backbone = backbone, batch_size = 64)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6VglyoT1hxi2",
        "outputId": "e2d604f9-e634-4994-9d12-d3226496c24c"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.77243, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.77243\n",
            "\n",
            "Epoch 3: val_loss improved from 2.77243 to 2.47702, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.47702\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.47702\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.47702\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.47702\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.47702\n",
            "\n",
            "Epoch 9: val_loss improved from 2.47702 to 1.48798, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.48798\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.48798\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.48798\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.48798\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.48798\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.48798\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.48798\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.48798\n",
            "12/12 [==============================] - 2s 53ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.77043, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.77043\n",
            "\n",
            "Epoch 3: val_loss improved from 2.77043 to 2.70491, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss improved from 2.70491 to 1.93527, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss improved from 1.93527 to 1.76554, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss improved from 1.76554 to 1.71572, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss improved from 1.71572 to 1.38118, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.38118\n",
            "\n",
            "Epoch 9: val_loss improved from 1.38118 to 1.17787, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.17787\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.17787\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.17787\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.17787\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.17787\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.17787\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.17787\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.17787\n",
            "12/12 [==============================] - 1s 56ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.49885, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss improved from 2.49885 to 2.19599, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.19599\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.19599\n",
            "\n",
            "Epoch 5: val_loss improved from 2.19599 to 2.08313, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.08313\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.08313\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.08313\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.08313\n",
            "\n",
            "Epoch 10: val_loss did not improve from 2.08313\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.08313\n",
            "\n",
            "Epoch 12: val_loss improved from 2.08313 to 1.56984, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 13: val_loss improved from 1.56984 to 1.17374, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.17374\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.17374\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.17374\n",
            "\n",
            "Epoch 17: val_loss improved from 1.17374 to 1.02214, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.02214\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.02214\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.02214\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.02214\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.02214\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.02214\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.02214\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.02214\n",
            "12/12 [==============================] - 1s 37ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.83001, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.83001\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.83001\n",
            "\n",
            "Epoch 4: val_loss improved from 2.83001 to 2.47619, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.47619\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.47619\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.47619\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.47619\n",
            "\n",
            "Epoch 9: val_loss improved from 2.47619 to 1.72864, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.72864\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.72864\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.72864\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.72864\n",
            "\n",
            "Epoch 14: val_loss improved from 1.72864 to 1.09051, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.09051\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.09051\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.09051\n",
            "\n",
            "Epoch 18: val_loss improved from 1.09051 to 1.01280, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.01280\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.01280\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.01280\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.01280\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.01280\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.01280\n",
            "\n",
            "Epoch 25: val_loss did not improve from 1.01280\n",
            "\n",
            "Epoch 26: val_loss did not improve from 1.01280\n",
            "12/12 [==============================] - 2s 68ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.64017, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.64017\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.64017\n",
            "\n",
            "Epoch 4: val_loss improved from 2.64017 to 2.41138, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.41138\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.41138\n",
            "\n",
            "Epoch 7: val_loss improved from 2.41138 to 1.79118, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.79118\n",
            "\n",
            "Epoch 9: val_loss improved from 1.79118 to 1.61540, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.61540\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.61540\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.61540\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.61540\n",
            "\n",
            "Epoch 14: val_loss improved from 1.61540 to 1.05516, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.05516\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.05516\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.05516\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.05516\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.05516\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.05516\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.05516\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.05516\n",
            "12/12 [==============================] - 1s 36ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.98256, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.98256\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.98256\n",
            "\n",
            "Epoch 4: val_loss improved from 2.98256 to 2.34807, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.34807\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.34807\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.34807\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.34807\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.34807\n",
            "\n",
            "Epoch 10: val_loss improved from 2.34807 to 2.16891, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss improved from 2.16891 to 2.16223, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 12: val_loss did not improve from 2.16223\n",
            "\n",
            "Epoch 13: val_loss improved from 2.16223 to 2.04365, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 14: val_loss did not improve from 2.04365\n",
            "\n",
            "Epoch 15: val_loss did not improve from 2.04365\n",
            "\n",
            "Epoch 16: val_loss did not improve from 2.04365\n",
            "\n",
            "Epoch 17: val_loss did not improve from 2.04365\n",
            "\n",
            "Epoch 18: val_loss did not improve from 2.04365\n",
            "\n",
            "Epoch 19: val_loss did not improve from 2.04365\n",
            "\n",
            "Epoch 20: val_loss did not improve from 2.04365\n",
            "\n",
            "Epoch 21: val_loss did not improve from 2.04365\n",
            "12/12 [==============================] - 1s 54ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.82911, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.82911\n",
            "\n",
            "Epoch 3: val_loss improved from 2.82911 to 2.37972, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.37972\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.37972\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.37972\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.37972\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.37972\n",
            "\n",
            "Epoch 9: val_loss did not improve from 2.37972\n",
            "\n",
            "Epoch 10: val_loss improved from 2.37972 to 2.18735, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 11: val_loss did not improve from 2.18735\n",
            "\n",
            "Epoch 12: val_loss improved from 2.18735 to 1.91369, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.91369\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.91369\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.91369\n",
            "\n",
            "Epoch 16: val_loss improved from 1.91369 to 1.71287, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.71287\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.71287\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.71287\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.71287\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.71287\n",
            "\n",
            "Epoch 22: val_loss did not improve from 1.71287\n",
            "\n",
            "Epoch 23: val_loss did not improve from 1.71287\n",
            "\n",
            "Epoch 24: val_loss did not improve from 1.71287\n",
            "12/12 [==============================] - 1s 37ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.71171, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.71171\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.71171\n",
            "\n",
            "Epoch 4: val_loss did not improve from 2.71171\n",
            "\n",
            "Epoch 5: val_loss improved from 2.71171 to 2.56257, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.56257\n",
            "\n",
            "Epoch 7: val_loss improved from 2.56257 to 1.74639, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.74639\n",
            "\n",
            "Epoch 9: val_loss did not improve from 1.74639\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.74639\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.74639\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.74639\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.74639\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.74639\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.74639\n",
            "12/12 [==============================] - 1s 47ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.68144, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.68144\n",
            "\n",
            "Epoch 3: val_loss improved from 2.68144 to 1.98614, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 4: val_loss did not improve from 1.98614\n",
            "\n",
            "Epoch 5: val_loss did not improve from 1.98614\n",
            "\n",
            "Epoch 6: val_loss improved from 1.98614 to 1.75653, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 7: val_loss did not improve from 1.75653\n",
            "\n",
            "Epoch 8: val_loss did not improve from 1.75653\n",
            "\n",
            "Epoch 9: val_loss improved from 1.75653 to 1.64238, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.64238\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.64238\n",
            "\n",
            "Epoch 12: val_loss improved from 1.64238 to 1.20993, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 13: val_loss improved from 1.20993 to 1.19403, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.19403\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.19403\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.19403\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.19403\n",
            "\n",
            "Epoch 18: val_loss did not improve from 1.19403\n",
            "\n",
            "Epoch 19: val_loss did not improve from 1.19403\n",
            "\n",
            "Epoch 20: val_loss did not improve from 1.19403\n",
            "\n",
            "Epoch 21: val_loss did not improve from 1.19403\n",
            "12/12 [==============================] - 1s 35ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cleared old weights\n",
            "\n",
            "Epoch 1: val_loss improved from inf to 2.71213, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Epoch 2: val_loss did not improve from 2.71213\n",
            "\n",
            "Epoch 3: val_loss did not improve from 2.71213\n",
            "\n",
            "Epoch 4: val_loss improved from 2.71213 to 2.66102, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 5: val_loss did not improve from 2.66102\n",
            "\n",
            "Epoch 6: val_loss did not improve from 2.66102\n",
            "\n",
            "Epoch 7: val_loss did not improve from 2.66102\n",
            "\n",
            "Epoch 8: val_loss did not improve from 2.66102\n",
            "\n",
            "Epoch 9: val_loss improved from 2.66102 to 1.47490, saving model to /content/gdrive/MyDrive/Colab Notebooks/data/m.h5\n",
            "\n",
            "Epoch 10: val_loss did not improve from 1.47490\n",
            "\n",
            "Epoch 11: val_loss did not improve from 1.47490\n",
            "\n",
            "Epoch 12: val_loss did not improve from 1.47490\n",
            "\n",
            "Epoch 13: val_loss did not improve from 1.47490\n",
            "\n",
            "Epoch 14: val_loss did not improve from 1.47490\n",
            "\n",
            "Epoch 15: val_loss did not improve from 1.47490\n",
            "\n",
            "Epoch 16: val_loss did not improve from 1.47490\n",
            "\n",
            "Epoch 17: val_loss did not improve from 1.47490\n",
            "12/12 [==============================] - 1s 35ms/step\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-10-0ddd3b46834f>:34: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  test_link[name] = pd.DataFrame(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "done in 2215.77\n"
          ]
        }
      ],
      "source": [
        "run_grouped_experiments(['dist', 'mixup', 'room', 'spectrum'],\"audio_only_nd\",backbone = backbone, batch_size = 64)"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "machine_shape": "hm",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.9.6"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}