Second_attempt.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Second_attempt",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "TPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/pren1/VAST/blob/master/Second_attempt.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "czEROTY595JP",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# %tensorflow_version 1.x"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "twI28paZ7ETZ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "outputId": "b78d378b-9d7d-40f7-b502-bbdd53d9b244"
      },
      "source": [
        "!rm -rf VAST/\n",
        "!git clone https://github.com/pren1/VAST.git\n",
        "!pip3 install soundfile\n",
        "!pip3 install librosa"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Cloning into 'VAST'...\n",
            "remote: Enumerating objects: 180, done.\u001b[K\n",
            "remote: Counting objects: 100% (180/180), done.\u001b[K\n",
            "remote: Compressing objects: 100% (145/145), done.\u001b[K\n",
            "remote: Total 180 (delta 84), reused 70 (delta 26), pack-reused 0\u001b[K\n",
            "Receiving objects: 100% (180/180), 64.12 MiB | 10.80 MiB/s, done.\n",
            "Resolving deltas: 100% (84/84), done.\n",
            "Collecting soundfile\n",
            "  Downloading https://files.pythonhosted.org/packages/eb/f2/3cbbbf3b96fb9fa91582c438b574cff3f45b29c772f94c400e2c99ef5db9/SoundFile-0.10.3.post1-py2.py3-none-any.whl\n",
            "Requirement already satisfied: cffi>=1.0 in /usr/local/lib/python3.6/dist-packages (from soundfile) (1.14.0)\n",
            "Requirement already satisfied: pycparser in /usr/local/lib/python3.6/dist-packages (from cffi>=1.0->soundfile) (2.20)\n",
            "Installing collected packages: soundfile\n",
            "Successfully installed soundfile-0.10.3.post1\n",
            "Requirement already satisfied: librosa in /usr/local/lib/python3.6/dist-packages (0.6.3)\n",
            "Requirement already satisfied: joblib>=0.12 in /usr/local/lib/python3.6/dist-packages (from librosa) (0.15.1)\n",
            "Requirement already satisfied: audioread>=2.0.0 in /usr/local/lib/python3.6/dist-packages (from librosa) (2.1.8)\n",
            "Requirement already satisfied: scipy>=1.0.0 in /usr/local/lib/python3.6/dist-packages (from librosa) (1.4.1)\n",
            "Requirement already satisfied: numpy>=1.8.0 in /usr/local/lib/python3.6/dist-packages (from librosa) (1.18.5)\n",
            "Requirement already satisfied: scikit-learn!=0.19.0,>=0.14.0 in /usr/local/lib/python3.6/dist-packages (from librosa) (0.22.2.post1)\n",
            "Requirement already satisfied: six>=1.3 in /usr/local/lib/python3.6/dist-packages (from librosa) (1.12.0)\n",
            "Requirement already satisfied: decorator>=3.0.0 in /usr/local/lib/python3.6/dist-packages (from librosa) (4.4.2)\n",
            "Requirement already satisfied: resampy>=0.2.0 in /usr/local/lib/python3.6/dist-packages (from librosa) (0.2.2)\n",
            "Requirement already satisfied: numba>=0.38.0 in /usr/local/lib/python3.6/dist-packages (from librosa) (0.48.0)\n",
            "Requirement already satisfied: llvmlite<0.32.0,>=0.31.0dev0 in /usr/local/lib/python3.6/dist-packages (from numba>=0.38.0->librosa) (0.31.0)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.6/dist-packages (from numba>=0.38.0->librosa) (47.3.1)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bAbJHx787OHV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from VAST.util import *\n",
        "import pdb\n",
        "import pprint\n",
        "from tqdm import tqdm\n",
        "import numpy as np\n",
        "import IPython\n",
        "import soundfile as sf\n",
        "import matplotlib.pyplot as plt\n",
        "import librosa\n",
        "import os\n",
        "from tqdm import trange"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CAksJZUD7OWN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 128
        },
        "outputId": "14782ec7-90c0-49ff-d17c-291ce5ba27ff"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/gdrive')"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/gdrive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IfYr7P78cWiU",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# adapted from Keith Ito's tacotron implementation\n",
        "# https://github.com/keithito/tacotron/blob/master/util/audio.py\n",
        "\n",
        "import librosa\n",
        "import numpy as np\n",
        "\n",
        "\n",
        "class Audio():\n",
        "    def __init__(self):\n",
        "        self.ref_level_db = 20.0\n",
        "        self.n_fft = 1200\n",
        "        self.hop_length = 160\n",
        "        self.win_length = 400\n",
        "        self.min_level_db = -100.0\n",
        "\n",
        "    def wav2spec(self, y):\n",
        "       \n",
        "    \n",
        "        D = self.stft(y)\n",
        "        S = self.amp_to_db(np.abs(D)) - self.ref_level_db\n",
        "        S, D = self.normalize(S), np.angle(D)\n",
        "        S, D = S.T, D.T # to make [time, freq]\n",
        "        return S, D\n",
        "\n",
        "    def spec2wav(self, spectrogram, phase):\n",
        "        spectrogram, phase = spectrogram.T, phase.T\n",
        "        # used during inference only\n",
        "        # spectrogram: enhanced output\n",
        "        # phase: use noisy input's phase, so no GLA is required\n",
        "        S = self.db_to_amp(self.denormalize(spectrogram) + self.ref_level_db)\n",
        "        return self.istft(S, phase)\n",
        "\n",
        "    def stft(self, y):\n",
        "        return librosa.stft(y=y, n_fft=self.n_fft,\n",
        "                            hop_length=self.hop_length,\n",
        "                            win_length=self.win_length)\n",
        "\n",
        "    def istft(self, mag, phase):\n",
        "        stft_matrix = mag * np.exp(1j*phase)\n",
        "        return librosa.istft(stft_matrix,\n",
        "                             hop_length=self.hop_length,\n",
        "                             win_length=self.win_length)\n",
        "\n",
        "    def amp_to_db(self, x):\n",
        "        return 20.0 * np.log10(np.maximum(1e-5, x))\n",
        "\n",
        "    def db_to_amp(self, x):\n",
        "        return np.power(10.0, x * 0.05)\n",
        "\n",
        "    def normalize(self, S):\n",
        "        return np.clip(S / -self.min_level_db, -1.0, 0.0) + 1.0\n",
        "\n",
        "    def denormalize(self, S):\n",
        "        return (np.clip(S, 0.0, 1.0) - 1.0) * -self.min_level_db\n",
        "\n",
        "audio = Audio()"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vo7t986-nHjx",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "embedding_matrix = np.asarray([\n",
        "         [-0.8265083 , -1.6049656 ,  1.5085448 ,  0.6589694 ,  2.1224942 ,\n",
        "         2.3647609 ,  0.11735925,  1.1543381 ,  0.99229884,  1.4278704 ,\n",
        "         0.5712279 , -1.5206277 ,  0.30854183, -2.6490169 ,  1.3902934 ,\n",
        "        -0.23975304,  0.94975144, -2.0093975 ,  0.17497058, -1.0554106 ,\n",
        "        -2.172874  , -1.5222347 ,  0.563387  ,  2.239781  , -0.96120906,\n",
        "        -0.42689592,  2.4739625 , -2.362247  , -0.98188835,  0.16969077,\n",
        "        -0.6247616 ,  0.04511972],\n",
        "        [ 0.47638106,  1.8903663 , -2.9872196 ,  2.3670852 , -2.4623342 ,\n",
        "         2.220159  ,  0.23939389,  1.5010741 , -0.7784126 , -0.54998654,\n",
        "        -2.0378437 , -0.07159348,  0.3601075 ,  1.7380389 ,  2.3474126 ,\n",
        "        -1.0004979 ,  0.47312295, -0.88486624,  0.18921344, -0.43055353,\n",
        "         0.38867134,  0.00616272, -3.2682106 , -2.6373188 ,  2.909382  ,\n",
        "         0.7305331 , -1.1071701 ,  0.9580342 , -1.1098799 , -1.7031851 ,\n",
        "         0.00362744,  2.3521018 ]\n",
        "         ])"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YJ7EoJwKnICJ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "bd429be7-76ab-492d-d742-79114fff7e4a"
      },
      "source": [
        "embedding_matrix.shape"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(2, 32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LwUWhLBm7VN_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import math\n",
        "import tensorflow as tf\n",
        "\n",
        "def trunk_network(conv_size = 64, LSTM_size = 400, input_size = (151, 601), dropout_rate = 0.0):\n",
        "    inputs = tf.keras.layers.Input(input_size)\n",
        "\n",
        "    speaker_embeddings = tf.keras.Input(shape=(1, 32), dtype=tf.float32)\n",
        "    \n",
        "    x = tf.keras.layers.Reshape((input_size[0],  input_size[1], 1))(inputs)\n",
        "    # cnn1\n",
        "    x = tf.keras.layers.ZeroPadding2D(padding=((0, 0), (3, 3)))(x)\n",
        "    x = tf.keras.layers.Conv2D(conv_size, (1, 7))(x)\n",
        "    x = tf.keras.layers.BatchNormalization()(x)\n",
        "    x = tf.keras.layers.ReLU()(x)\n",
        "    x = tf.keras.layers.Dropout(dropout_rate)(x)\n",
        "\n",
        "    # cnn2\n",
        "    x = tf.keras.layers.ZeroPadding2D(padding=((3, 3), (0, 0)))(x)\n",
        "    x = tf.keras.layers.Conv2D(conv_size, (7, 1))(x)\n",
        "    x = tf.keras.layers.BatchNormalization()(x)\n",
        "    x = tf.keras.layers.ReLU()(x)\n",
        "    x = tf.keras.layers.Dropout(dropout_rate)(x)\n",
        "\n",
        "    # cnn3\n",
        "    x = tf.keras.layers.ZeroPadding2D(padding=((2, 2), (2, 2)))(x)\n",
        "    x = tf.keras.layers.Conv2D(conv_size, (5, 5))(x)\n",
        "    x = tf.keras.layers.BatchNormalization()(x)\n",
        "    x = tf.keras.layers.ReLU()(x)\n",
        "    x = tf.keras.layers.Dropout(dropout_rate)(x)\n",
        "\n",
        "    # # cnn1\n",
        "    # x = tf.keras.layers.ZeroPadding2D(padding=((0, 0), (3, 3)))(x)\n",
        "    # x = tf.keras.layers.Conv2D(conv_size, (1, 7))(x)\n",
        "    # x = tf.keras.layers.BatchNormalization()(x)\n",
        "    # x = tf.keras.layers.ReLU()(x)\n",
        "    # x = tf.keras.layers.Dropout(dropout_rate)(x)\n",
        "\n",
        "    # # cnn2\n",
        "    # x = tf.keras.layers.ZeroPadding2D(padding=((3, 3), (0, 0)))(x)\n",
        "    # x = tf.keras.layers.Conv2D(conv_size, (7, 1))(x)\n",
        "    # x = tf.keras.layers.BatchNormalization()(x)\n",
        "    # x = tf.keras.layers.ReLU()(x)\n",
        "    # x = tf.keras.layers.Dropout(dropout_rate)(x)\n",
        "\n",
        "    # cnn3\n",
        "    x = tf.keras.layers.ZeroPadding2D(padding=((2, 2), (2, 2)))(x)\n",
        "    x = tf.keras.layers.Conv2D(conv_size, (5, 5))(x)\n",
        "    x = tf.keras.layers.BatchNormalization()(x)\n",
        "    x = tf.keras.layers.ReLU()(x)\n",
        "    x = tf.keras.layers.Dropout(dropout_rate)(x)\n",
        "\n",
        "    # cnn3\n",
        "    x = tf.keras.layers.ZeroPadding2D(padding=((2, 2), (2, 2)))(x)\n",
        "    x = tf.keras.layers.Conv2D(conv_size, (5, 5))(x)\n",
        "    x = tf.keras.layers.BatchNormalization()(x)\n",
        "    x = tf.keras.layers.ReLU()(x)\n",
        "    x = tf.keras.layers.Dropout(dropout_rate)(x)\n",
        "\n",
        "    # cnn3\n",
        "    x = tf.keras.layers.ZeroPadding2D(padding=((2, 2), (2, 2)))(x)\n",
        "    x = tf.keras.layers.Conv2D(conv_size, (5, 5))(x)\n",
        "    x = tf.keras.layers.BatchNormalization()(x)\n",
        "    x = tf.keras.layers.ReLU()(x)\n",
        "    x = tf.keras.layers.Dropout(dropout_rate)(x)\n",
        "\n",
        "    # # # cnn4\n",
        "    # x = tf.keras.layers.ZeroPadding2D(padding=((4, 4), (2, 2)))(x)\n",
        "    # x = tf.keras.layers.Conv2D(conv_size, (5, 5), dilation_rate=(2, 1))(x)\n",
        "    # x = tf.keras.layers.BatchNormalization()(x)\n",
        "    # x = tf.keras.layers.ReLU()(x)\n",
        "    # pdb.set_trace()\n",
        "\n",
        "    # # cnn5\n",
        "    # x = tf.keras.layers.ZeroPadding2D(padding=((8, 8), (2, 2)))(x)     \n",
        "    # x = tf.keras.layers.Conv2D(conv_size, (5, 5), dilation_rate=(4, 1))(x)       \n",
        "    # x = tf.keras.layers.BatchNormalization()(x)\n",
        "    # x = tf.keras.layers.ReLU()(x)\n",
        "\n",
        "    # # cnn6\n",
        "    # x = tf.keras.layers.ZeroPadding2D(padding=((16, 16), (2, 2)))(x)                 \n",
        "    # x = tf.keras.layers.Conv2D(conv_size, (5, 5), dilation_rate=(8, 1))(x) \n",
        "    # x = tf.keras.layers.BatchNormalization()(x)\n",
        "    # x = tf.keras.layers.ReLU()(x)\n",
        "\n",
        "    # # cnn7 TensorShape([None, 51, 601, conv_size])\n",
        "    # x = tf.keras.layers.ZeroPadding2D(padding=((32, 32), (2, 2)))(x)\n",
        "    # x = tf.keras.layers.Conv2D(conv_size, (5, 5), dilation_rate=(16, 1))(x)     \n",
        "    # x = tf.keras.layers.BatchNormalization()(x)\n",
        "    # x = tf.keras.layers.ReLU()(x)\n",
        "\n",
        "    # cnn8 TensorShape([None, 51, 601, 8])\n",
        "    x = tf.keras.layers.Conv2D(8, (1, 1))(x)     \n",
        "    x = tf.keras.layers.BatchNormalization()(x)\n",
        "    x = tf.keras.layers.ReLU()(x)\n",
        "    x = tf.keras.layers.Dropout(dropout_rate)(x)\n",
        "\n",
        "    'Reshape x shape TensorShape([None, 51, 4808])'\n",
        "    # x = tf.keras.layers.Reshape((tf.keras.backend.int_shape(x)[1], tf.keras.backend.int_shape(x)[2]*tf.keras.backend.int_shape(x)[3]))(x)\n",
        "    x = tf.keras.layers.Reshape((151, 4808))(x)\n",
        "    repeat_embedding_time_wise = tf.tile(speaker_embeddings, [1, 151, 1])\n",
        "    \n",
        "    x = tf.keras.layers.concatenate([x, repeat_embedding_time_wise],axis=-1)\n",
        "    \n",
        "    attention_handle = x\n",
        "  \n",
        "    # LSTM 1\n",
        "    x = tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(LSTM_size, return_sequences=True)\n",
        "                            )(x)  # [b_s, seq_len, vec_dim]\n",
        "    x = tf.keras.layers.Dropout(dropout_rate)(x)                        \n",
        "\n",
        "    # # LSTM 2 TensorShape([None, 51, 800])                                           \n",
        "    x = tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(LSTM_size, return_sequences=True)\n",
        "                            )(x)  # [b_s, seq_len, vec_dim]\n",
        "\n",
        "    'build_attention: Do not forget to change the dim here if you change the LSTM dim...'\n",
        "    attention_handle = tf.keras.layers.TimeDistributed(tf.keras.layers.Dense(LSTM_size * 2))(attention_handle)\n",
        "    attention = tf.keras.layers.Dot(axes=[2, 2])([x, attention_handle])\n",
        "    attention = tf.keras.layers.Activation('softmax', name='attention')(attention)\n",
        "    context = tf.keras.layers.Dot(axes=[2, 1])([attention, attention_handle])\n",
        "    'concatenate it with x'\n",
        "    x = tf.keras.layers.concatenate([x, context],axis=-1)\n",
        "\n",
        "    x = tf.keras.layers.ReLU()(x)\n",
        "    x = tf.keras.layers.Dropout(dropout_rate)(x) \n",
        "    # Dense 1\n",
        "    x = tf.keras.layers.TimeDistributed(tf.keras.layers.Dense(input_size[1]))(x)\n",
        "    x = tf.keras.layers.ReLU()(x)\n",
        "    x = tf.keras.layers.Dropout(dropout_rate)(x)\n",
        "    \n",
        "    # pdb.set_trace()\n",
        "    # # Try to add attention..\n",
        "    # '1. combine x and attention_handler'\n",
        "    # dot_1 = tf.keras.layers.Dot(axes=[1, 1])([x, attention_handle])\n",
        "    # attention_weights = tf.keras.layers.Activation('softmax', name='attention')(dot_1)\n",
        "    # '2. assigned the weights to the previous output'\n",
        "    # attentioned_res = tf.keras.layers.Dot(axes=[2, 2])([attention_handle, attention_weights])\n",
        "    # 'concatenate it with x'\n",
        "    # x = tf.keras.layers.concatenate([x, attentioned_res],axis=-1)\n",
        "    \n",
        "    # Dense 2\n",
        "    mask = tf.keras.layers.TimeDistributed(tf.keras.layers.Dense(input_size[1], activation='sigmoid'))(x)\n",
        "    output = tf.keras.layers.Multiply()([inputs, mask])\n",
        "    # print(output.shape)\n",
        "    model = tf.keras.Model(inputs=[inputs, speaker_embeddings], outputs=[output])\n",
        "    tf.keras.utils.plot_model(model, show_shapes=True, to_file='model.png')\n",
        "    return model\n",
        "\n",
        "'learning rate decay'\n",
        "def step_decay(epoch):\n",
        "    initial_lrate = 0.0001\n",
        "    drop = 0.7\n",
        "    epochs_drop = 20.0\n",
        "    lrate = initial_lrate * math.pow(drop, math.floor((1 + epoch) / epochs_drop))\n",
        "    return lrate   \n",
        "# trunk_network()"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QBwhSxHcVS1E",
        "colab_type": "text"
      },
      "source": [
        "### Prepare data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9ehDMj2ac7S-",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 745
        },
        "outputId": "98f57ad1-e810-4ec8-b0ab-750eb2cda5ea"
      },
      "source": [
        "resolver = tf.distribute.cluster_resolver.TPUClusterResolver(tpu='grpc://' + os.environ['COLAB_TPU_ADDR'])\n",
        "tf.config.experimental_connect_to_cluster(resolver)\n",
        "# This is the TPU initialization code that has to be at the beginning.\n",
        "tf.tpu.experimental.initialize_tpu_system(resolver)\n",
        "strategy = tf.distribute.experimental.TPUStrategy(resolver)\n",
        "with strategy.scope():\n",
        "  model = trunk_network(conv_size = 64, LSTM_size = 256)\n",
        "  model.compile(optimizer=tf.keras.optimizers.Adam(0.001), loss=['mse'])\n",
        "  lrate = tf.keras.callbacks.LearningRateScheduler(step_decay)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Initializing the TPU system: grpc://10.5.207.74:8470\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Initializing the TPU system: grpc://10.5.207.74:8470\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Clearing out eager caches\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Clearing out eager caches\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Finished initializing TPU system.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Finished initializing TPU system.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Found TPU system:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Found TPU system:\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Num TPU Cores: 8\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Num TPU Cores: 8\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Num TPU Workers: 1\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Num TPU Workers: 1\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Num TPU Cores Per Worker: 8\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Num TPU Cores Per Worker: 8\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:localhost/replica:0/task:0/device:CPU:0, CPU, 0, 0)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:localhost/replica:0/task:0/device:CPU:0, CPU, 0, 0)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:localhost/replica:0/task:0/device:XLA_CPU:0, XLA_CPU, 0, 0)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:localhost/replica:0/task:0/device:XLA_CPU:0, XLA_CPU, 0, 0)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:CPU:0, CPU, 0, 0)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:CPU:0, CPU, 0, 0)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:0, TPU, 0, 0)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:0, TPU, 0, 0)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:1, TPU, 0, 0)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:1, TPU, 0, 0)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:2, TPU, 0, 0)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:2, TPU, 0, 0)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:3, TPU, 0, 0)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:3, TPU, 0, 0)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:4, TPU, 0, 0)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:4, TPU, 0, 0)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:5, TPU, 0, 0)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:5, TPU, 0, 0)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:6, TPU, 0, 0)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:6, TPU, 0, 0)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:7, TPU, 0, 0)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU:7, TPU, 0, 0)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU_SYSTEM:0, TPU_SYSTEM, 0, 0)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:TPU_SYSTEM:0, TPU_SYSTEM, 0, 0)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:XLA_CPU:0, XLA_CPU, 0, 0)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:*** Available Device: _DeviceAttributes(/job:worker/replica:0/task:0/device:XLA_CPU:0, XLA_CPU, 0, 0)\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ylpzMlb5jQL2",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "13cdc662-0c7b-454c-dbd0-333c5929dc34"
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"model\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_1 (InputLayer)            [(None, 151, 601)]   0                                            \n",
            "__________________________________________________________________________________________________\n",
            "reshape (Reshape)               (None, 151, 601, 1)  0           input_1[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d (ZeroPadding2D)  (None, 151, 607, 1)  0           reshape[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "conv2d (Conv2D)                 (None, 151, 601, 64) 512         zero_padding2d[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization (BatchNorma (None, 151, 601, 64) 256         conv2d[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "re_lu (ReLU)                    (None, 151, 601, 64) 0           batch_normalization[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "dropout (Dropout)               (None, 151, 601, 64) 0           re_lu[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_1 (ZeroPadding2D (None, 157, 601, 64) 0           dropout[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_1 (Conv2D)               (None, 151, 601, 64) 28736       zero_padding2d_1[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_1 (BatchNor (None, 151, 601, 64) 256         conv2d_1[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "re_lu_1 (ReLU)                  (None, 151, 601, 64) 0           batch_normalization_1[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_1 (Dropout)             (None, 151, 601, 64) 0           re_lu_1[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_2 (ZeroPadding2D (None, 155, 605, 64) 0           dropout_1[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_2 (Conv2D)               (None, 151, 601, 64) 102464      zero_padding2d_2[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_2 (BatchNor (None, 151, 601, 64) 256         conv2d_2[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "re_lu_2 (ReLU)                  (None, 151, 601, 64) 0           batch_normalization_2[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_2 (Dropout)             (None, 151, 601, 64) 0           re_lu_2[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_3 (ZeroPadding2D (None, 155, 605, 64) 0           dropout_2[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_3 (Conv2D)               (None, 151, 601, 64) 102464      zero_padding2d_3[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_3 (BatchNor (None, 151, 601, 64) 256         conv2d_3[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "re_lu_3 (ReLU)                  (None, 151, 601, 64) 0           batch_normalization_3[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_3 (Dropout)             (None, 151, 601, 64) 0           re_lu_3[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_4 (ZeroPadding2D (None, 155, 605, 64) 0           dropout_3[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_4 (Conv2D)               (None, 151, 601, 64) 102464      zero_padding2d_4[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_4 (BatchNor (None, 151, 601, 64) 256         conv2d_4[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "re_lu_4 (ReLU)                  (None, 151, 601, 64) 0           batch_normalization_4[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_4 (Dropout)             (None, 151, 601, 64) 0           re_lu_4[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_5 (ZeroPadding2D (None, 155, 605, 64) 0           dropout_4[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_5 (Conv2D)               (None, 151, 601, 64) 102464      zero_padding2d_5[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_5 (BatchNor (None, 151, 601, 64) 256         conv2d_5[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "re_lu_5 (ReLU)                  (None, 151, 601, 64) 0           batch_normalization_5[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_5 (Dropout)             (None, 151, 601, 64) 0           re_lu_5[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_6 (Conv2D)               (None, 151, 601, 8)  520         dropout_5[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_6 (BatchNor (None, 151, 601, 8)  32          conv2d_6[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "re_lu_6 (ReLU)                  (None, 151, 601, 8)  0           batch_normalization_6[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_6 (Dropout)             (None, 151, 601, 8)  0           re_lu_6[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "input_2 (InputLayer)            [(None, 1, 32)]      0                                            \n",
            "__________________________________________________________________________________________________\n",
            "reshape_1 (Reshape)             (None, 151, 4808)    0           dropout_6[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "tf_op_layer_Tile (TensorFlowOpL [(None, 151, 32)]    0           input_2[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "concatenate (Concatenate)       (None, 151, 4840)    0           reshape_1[0][0]                  \n",
            "                                                                 tf_op_layer_Tile[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "bidirectional (Bidirectional)   (None, 151, 512)     10438656    concatenate[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "dropout_7 (Dropout)             (None, 151, 512)     0           bidirectional[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "bidirectional_1 (Bidirectional) (None, 151, 512)     1574912     dropout_7[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "time_distributed (TimeDistribut (None, 151, 512)     2478592     concatenate[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "dot (Dot)                       (None, 151, 151)     0           bidirectional_1[0][0]            \n",
            "                                                                 time_distributed[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "attention (Activation)          (None, 151, 151)     0           dot[0][0]                        \n",
            "__________________________________________________________________________________________________\n",
            "dot_1 (Dot)                     (None, 151, 512)     0           attention[0][0]                  \n",
            "                                                                 time_distributed[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_1 (Concatenate)     (None, 151, 1024)    0           bidirectional_1[0][0]            \n",
            "                                                                 dot_1[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "re_lu_7 (ReLU)                  (None, 151, 1024)    0           concatenate_1[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "dropout_8 (Dropout)             (None, 151, 1024)    0           re_lu_7[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "time_distributed_1 (TimeDistrib (None, 151, 601)     616025      dropout_8[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "re_lu_8 (ReLU)                  (None, 151, 601)     0           time_distributed_1[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "dropout_9 (Dropout)             (None, 151, 601)     0           re_lu_8[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "time_distributed_2 (TimeDistrib (None, 151, 601)     361802      dropout_9[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "multiply (Multiply)             (None, 151, 601)     0           input_1[0][0]                    \n",
            "                                                                 time_distributed_2[0][0]         \n",
            "==================================================================================================\n",
            "Total params: 15,911,179\n",
            "Trainable params: 15,910,395\n",
            "Non-trainable params: 784\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Rpb-tUHJ8QqT",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# The batch iterator\n",
        "class BatchIterator:\n",
        "    def __init__(self, data, label, spk_label, batch_size, is_shuffle):\n",
        "        data = np.asarray(data)\n",
        "        label = np.asarray(label)\n",
        "        self.spk_label = np.asarray(spk_label)\n",
        "        self.data = data\n",
        "        self.label = label\n",
        "        self.batch_size = batch_size\n",
        "        self.cursor = 0\n",
        "        self.size = data.shape[0]\n",
        "        self.order = np.arange(self.size)\n",
        "        self.shuffled = False\n",
        "        if is_shuffle:\n",
        "            self.shuffle()\n",
        "\n",
        "    def shuffle(self):\n",
        "        np.random.shuffle(self.order)\n",
        "        # print \"Batch shuffled\"\n",
        "        self.cursor = 0\n",
        "\n",
        "    def next(self):\n",
        "        if self.shuffled==True:\n",
        "            self.shuffled = False\n",
        "        if self.cursor + self.batch_size > self.size:\n",
        "            self.shuffle()\n",
        "            self.shuffled = True\n",
        "        idx = self.order[self.cursor:self.cursor + self.batch_size]\n",
        "        self.cursor += self.batch_size\n",
        "        return self.data[idx], self.label[idx], self.spk_label[idx]"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cMvHoLTsSfCe",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def evaluate_results(model, X_test, y_test, test_label):\n",
        "  test_res = model.predict([X_test, test_label])\n",
        "  mse = (np.square(test_res - y_test)).mean(axis=None)\n",
        "  return mse"
      ],
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R9pK3R-bUnnE",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.load_weights(f'gdrive/My Drive/Spitter_VTB/205_four_speaker_split.h5')\n",
        "# model.load_weights(f'gdrive/My Drive/Spitter_VTB/test_speaker_400.h5')\n",
        "# model.load_weights(f'gdrive/My Drive/Spitter_VTB/test_speaker.h5')\n",
        "# model.load_weights(f'gdrive/My Drive/Spitter_VTB/two_test_speakers.h5')\n",
        "# model.load_weights(f'gdrive/My Drive/Spitter_VTB/my_test_speakers.h5')\n",
        "# model.load_weights(f'gdrive/My Drive/Spitter_VTB/att1_test_speakers.h5')\n",
        "# model.load_weights(f'gdrive/My Drive/Spitter_VTB/real_atten_speakers_3.h5')\n",
        "# model.load_weights(f'gdrive/My Drive/Spitter_VTB/real_atten_speakers_cool.h5')"
      ],
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MvjLjF46fS3H",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# model.save('voice_splitter.h5')"
      ],
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uNZAbQF3zlt2",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "7757e026-6698-4053-a4b5-f5083b68903a"
      },
      "source": [
        "# Baby sitter: at 128 iter, changed lr from 0.001 to 1e-4\n",
        "iterations = 240\n",
        "# iterations = 32\n",
        "X_test = load_data_array_from_npy(f\"gdrive/My Drive/More_attempt/scale_test_case/cycle_mixed_mag_list.npy\")\n",
        "y_test = load_data_array_from_npy(f\"gdrive/My Drive/More_attempt/scale_test_case/cycle_target_mag_list.npy\")\n",
        "test_label = load_data_array_from_npy(f\"gdrive/My Drive/More_attempt/scale_test_case/cycle_speaker_label.npy\")\n",
        "test_label = np.expand_dims(np.matmul(np.eye(2)[np.asarray(test_label)], embedding_matrix), axis=1)\n",
        "\n",
        "data_counter = 0\n",
        "with trange(iterations) as t:\n",
        "  for iteration in t:\n",
        "    if True:\n",
        "      'load X_train'\n",
        "      # target_index = (data_counter + 28) % 60\n",
        "      # target_index = (data_counter) % 60\n",
        "      # target_index = (data_counter) % 10 + 60\n",
        "      target_index = (data_counter) % 60\n",
        "\n",
        "      X_train = load_data_array_from_npy(f\"gdrive/My Drive/More_attempt/scale_train_case/cycle_mixed_mag_list_{target_index}.npy\")\n",
        "      y_train = load_data_array_from_npy(f\"gdrive/My Drive/More_attempt/scale_train_case/cycle_target_mag_list_{target_index}.npy\")\n",
        "      train_label = load_data_array_from_npy(f\"gdrive/My Drive/More_attempt/scale_train_case/cycle_speaker_label{target_index}.npy\")\n",
        "      \n",
        "      train_label = np.expand_dims(np.matmul(np.eye(2)[np.asarray(train_label)], embedding_matrix), axis=1)\n",
        "\n",
        "      print(f\"New data loaded at {target_index}\")\n",
        "      data_counter += 1\n",
        "\n",
        "    model.fit(\n",
        "      x = [X_train, train_label], y = y_train,\n",
        "      # steps_per_epoch=67,\n",
        "      batch_size = 32,\n",
        "      epochs=2,\n",
        "      validation_data = ([X_test, test_label], y_test),\n",
        "      callbacks=[lrate]\n",
        "    )\n",
        "\n",
        "    if iteration % 5 == 0:\n",
        "      model.save_weights(f'gdrive/My Drive/Spitter_VTB/{iteration}_four_speaker_split.h5')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\r  0%|          | 0/240 [00:00<?, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 0\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 69s 733ms/step - loss: 0.0070 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0068 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  0%|          | 1/240 [03:10<12:38:55, 190.52s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 1\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0068 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0066 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  1%|          | 2/240 [05:26<11:30:23, 174.05s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 2\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 46s 484ms/step - loss: 0.0068 - val_loss: 0.0077 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 481ms/step - loss: 0.0066 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  1%|▏         | 3/240 [07:38<10:38:00, 161.52s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 3\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 483ms/step - loss: 0.0068 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0066 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  2%|▏         | 4/240 [09:57<10:09:01, 154.84s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 4\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0068 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 483ms/step - loss: 0.0066 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  2%|▏         | 5/240 [12:09<9:39:31, 147.96s/it] "
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 5\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0067 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0066 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  2%|▎         | 6/240 [14:22<9:19:28, 143.46s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 6\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0068 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0067 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  3%|▎         | 7/240 [16:41<9:11:51, 142.11s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 7\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0068 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0066 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  3%|▎         | 8/240 [18:53<8:58:15, 139.21s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 8\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0067 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0065 - val_loss: 0.0077 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  4%|▍         | 9/240 [21:07<8:49:31, 137.54s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 9\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0067 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0065 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  4%|▍         | 10/240 [23:21<8:43:24, 136.54s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 10\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0068 - val_loss: 0.0077 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 46s 495ms/step - loss: 0.0067 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  5%|▍         | 11/240 [25:36<8:38:44, 135.91s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 11\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0068 - val_loss: 0.0077 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0066 - val_loss: 0.0077 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  5%|▌         | 12/240 [27:50<8:34:29, 135.39s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 12\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 482ms/step - loss: 0.0067 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0065 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  5%|▌         | 13/240 [30:04<8:30:41, 134.98s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 13\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0068 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 46s 488ms/step - loss: 0.0066 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  6%|▌         | 14/240 [32:20<8:29:44, 135.33s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 14\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0066 - val_loss: 0.0077 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0064 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  6%|▋         | 15/240 [34:34<8:26:04, 134.95s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 15\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 481ms/step - loss: 0.0067 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0065 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  7%|▋         | 16/240 [36:49<8:24:09, 135.04s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 16\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 46s 484ms/step - loss: 0.0067 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0065 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  7%|▋         | 17/240 [39:05<8:22:14, 135.13s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 17\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0067 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0065 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  8%|▊         | 18/240 [41:19<8:18:51, 134.83s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 18\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0067 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0066 - val_loss: 0.0077 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  8%|▊         | 19/240 [43:32<8:14:51, 134.35s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 19\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 46s 494ms/step - loss: 0.0067 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0065 - val_loss: 0.0077 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  8%|▊         | 20/240 [45:49<8:15:52, 135.24s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 20\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 483ms/step - loss: 0.0070 - val_loss: 0.0078 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 483ms/step - loss: 0.0068 - val_loss: 0.0077 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  9%|▉         | 21/240 [48:06<8:15:37, 135.79s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 21\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0069 - val_loss: 0.0077 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0067 - val_loss: 0.0077 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r  9%|▉         | 22/240 [50:23<8:13:53, 135.94s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 22\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 481ms/step - loss: 0.0069 - val_loss: 0.0077 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0067 - val_loss: 0.0078 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 10%|▉         | 23/240 [52:46<8:19:50, 138.21s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 23\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0069 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0067 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 10%|█         | 24/240 [55:05<8:18:04, 138.35s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 24\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0068 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0066 - val_loss: 0.0077 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 10%|█         | 25/240 [57:21<8:13:25, 137.70s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 25\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 481ms/step - loss: 0.0071 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0069 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 11%|█         | 26/240 [59:41<8:13:05, 138.25s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 26\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0074 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0071 - val_loss: 0.0077 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 11%|█▏        | 27/240 [1:02:00<8:12:13, 138.66s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 27\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 481ms/step - loss: 0.0070 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0068 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 12%|█▏        | 28/240 [1:04:20<8:11:26, 139.09s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 28\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 47s 495ms/step - loss: 0.0071 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0069 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 12%|█▏        | 29/240 [1:06:41<8:10:48, 139.57s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 29\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 46s 484ms/step - loss: 0.0071 - val_loss: 0.0077 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0069 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 12%|█▎        | 30/240 [1:09:09<8:16:45, 141.93s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 30\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0072 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 481ms/step - loss: 0.0070 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 13%|█▎        | 31/240 [1:11:31<8:15:08, 142.15s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 31\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0072 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 482ms/step - loss: 0.0069 - val_loss: 0.0077 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 13%|█▎        | 32/240 [1:13:48<8:07:28, 140.62s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 32\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0070 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0068 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 14%|█▍        | 33/240 [1:16:10<8:06:23, 140.98s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 33\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0072 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0070 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 14%|█▍        | 34/240 [1:18:25<7:57:45, 139.15s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 34\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0072 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0069 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 15%|█▍        | 35/240 [1:20:44<7:55:34, 139.19s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 35\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0072 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0069 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 15%|█▌        | 36/240 [1:23:04<7:54:14, 139.48s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 36\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0072 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0070 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 15%|█▌        | 37/240 [1:25:25<7:52:59, 139.80s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 37\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 483ms/step - loss: 0.0072 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 481ms/step - loss: 0.0069 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 16%|█▌        | 38/240 [1:27:46<7:51:32, 140.06s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 38\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 482ms/step - loss: 0.0073 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 481ms/step - loss: 0.0070 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 16%|█▋        | 39/240 [1:30:06<7:49:24, 140.12s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 39\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0072 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 481ms/step - loss: 0.0069 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 17%|█▋        | 40/240 [1:32:26<7:47:10, 140.15s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 40\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 46s 485ms/step - loss: 0.0071 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 482ms/step - loss: 0.0068 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 17%|█▋        | 41/240 [1:34:45<7:43:46, 139.83s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 41\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0075 - val_loss: 0.0077 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0071 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 18%|█▊        | 42/240 [1:37:06<7:42:12, 140.06s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 42\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0073 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 46s 485ms/step - loss: 0.0070 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 18%|█▊        | 43/240 [1:39:24<7:37:49, 139.44s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 43\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0071 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 481ms/step - loss: 0.0069 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 18%|█▊        | 44/240 [1:41:41<7:33:18, 138.77s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 44\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 482ms/step - loss: 0.0072 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 481ms/step - loss: 0.0070 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 19%|█▉        | 45/240 [1:43:59<7:30:12, 138.52s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 45\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0072 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 484ms/step - loss: 0.0070 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 19%|█▉        | 46/240 [1:46:18<7:28:40, 138.76s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 46\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 481ms/step - loss: 0.0071 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0069 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 20%|█▉        | 47/240 [1:48:32<7:21:52, 137.37s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 47\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 481ms/step - loss: 0.0072 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0069 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 20%|██        | 48/240 [1:50:45<7:15:10, 135.99s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 48\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0072 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 477ms/step - loss: 0.0069 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 20%|██        | 49/240 [1:52:58<7:10:08, 135.12s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 49\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0072 - val_loss: 0.0077 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0070 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 21%|██        | 50/240 [1:55:11<7:05:18, 134.31s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 50\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0073 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0070 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 21%|██▏       | 51/240 [1:57:25<7:03:09, 134.33s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 51\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0072 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0070 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 22%|██▏       | 52/240 [1:59:39<7:00:15, 134.13s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 52\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0072 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0070 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 22%|██▏       | 53/240 [2:01:52<6:57:42, 134.02s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 53\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0074 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0071 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 22%|██▎       | 54/240 [2:04:07<6:55:51, 134.15s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 54\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 477ms/step - loss: 0.0072 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0070 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 23%|██▎       | 55/240 [2:06:23<6:55:43, 134.83s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 55\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 484ms/step - loss: 0.0072 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0070 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 23%|██▎       | 56/240 [2:08:40<6:54:54, 135.30s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 56\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0072 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0069 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 24%|██▍       | 57/240 [2:10:55<6:53:05, 135.44s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 57\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0072 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0069 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 24%|██▍       | 58/240 [2:13:08<6:47:49, 134.45s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 58\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 480ms/step - loss: 0.0072 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0070 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 25%|██▍       | 59/240 [2:15:24<6:47:25, 135.06s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 59\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0072 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0070 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 25%|██▌       | 60/240 [2:17:39<6:45:03, 135.02s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 0\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0069 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0067 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 25%|██▌       | 61/240 [2:20:03<6:50:31, 137.61s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 1\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0066 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0065 - val_loss: 0.0076 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 26%|██▌       | 62/240 [2:22:16<6:44:27, 136.33s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 2\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 478ms/step - loss: 0.0067 - val_loss: 0.0075 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 46s 492ms/step - loss: 0.0065 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 26%|██▋       | 63/240 [2:24:29<6:39:28, 135.42s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 3\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0067 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0065 - val_loss: 0.0075 - lr: 1.0000e-04\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r 27%|██▋       | 64/240 [2:26:42<6:34:38, 134.54s/it]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "New data loaded at 4\n",
            "Epoch 1/2\n",
            "94/94 [==============================] - 45s 479ms/step - loss: 0.0067 - val_loss: 0.0076 - lr: 1.0000e-04\n",
            "Epoch 2/2\n",
            "50/94 [==============>...............] - ETA: 15s - loss: 0.0065"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2S14xv46UIt4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# model.save_weights(f'gdrive/My Drive/Spitter_VTB/real_atten_speakers_cool.h5')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0z7Dddn8sKuH",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X_test = load_data_array_from_npy(f\"gdrive/My Drive/More_attempt/scale_val_case/cycle_mixed_mag_list.npy\")\n",
        "y_test = load_data_array_from_npy(f\"gdrive/My Drive/More_attempt/scale_val_case/cycle_target_mag_list.npy\")\n",
        "test_label = load_data_array_from_npy(f\"gdrive/My Drive/More_attempt/scale_val_case/cycle_speaker_label.npy\")\n",
        "test_label = np.expand_dims(np.matmul(np.eye(2)[np.asarray(test_label)], embedding_matrix), axis=1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y4AZ3NNTg93Z",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "test_res = model.predict([X_test, test_label])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bxCsPd1mhL88",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "test_res.shape"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iguo1isWfLSP",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "np.expand_dims(embedding_matrix[0, :], axis=0).shape"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y7FqYldIfNAa",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "test_label[0].shape"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bnlhIWbDV9Tq",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def reconstruct_wav(mag, phase, plt_name):\n",
        "  resulted_wav = audio.spec2wav(mag, phase)\n",
        "  plt.plot(resulted_wav, color = 'orange', label='resulted_wav')\n",
        "  plt.title(plt_name)\n",
        "  plt.show()\n",
        "  IPython.display.display(IPython.display.Audio(np.asarray(resulted_wav).T, rate=8000))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IFC_3ufx2UlN",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "test_phase = load_data_array_from_npy(f\"gdrive/My Drive/More_attempt/scale_val_case/cycle_mixed_phase_list.npy\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3eq64CegxobJ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "test = load_data_array_from_npy(f\"gdrive/My Drive/More_attempt/scale_val_case/cycle_speaker_label.npy\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ya_32l_FyZtz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "test.tolist()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZAq2AX3gy9ae",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "test_label[0]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QkLltiOMWlNO",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# for i in range(10, 20):\n",
        "#   print(test_label[i])\n",
        "#   reconstruct_wav(X_test[i], test_phase[i], \"Mixed voice\")\n",
        "#   reconstruct_wav(y_test[i], test_phase[i], \"ground truth\")\n",
        "#   reconstruct_wav(test_res[i], test_phase[i], \"prediction\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NIJeVD6BfaXA",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "label_0_embeddings = np.expand_dims(embedding_matrix[0, :], axis=0)\n",
        "label_1_embeddings = np.expand_dims(embedding_matrix[1, :], axis=0)\n",
        "\n",
        "for i in range(30, 40):\n",
        "  label_0_res = model.predict([np.expand_dims(X_test[i], axis=0), np.expand_dims(label_0_embeddings, axis=0)])\n",
        "  label_1_res = model.predict([np.expand_dims(X_test[i], axis=0), np.expand_dims(label_1_embeddings, axis=0)])\n",
        "\n",
        "  reconstruct_wav(X_test[i], test_phase[i], \"Mixed voice\")\n",
        "  reconstruct_wav(y_test[i], test_phase[i], \"ground truth\")\n",
        "  reconstruct_wav(label_0_res[0], test_phase[i], \"prediction: label 0\")\n",
        "  reconstruct_wav(label_1_res[0], test_phase[i], \"prediction: label 1\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qdmGBq8ctKDB",
        "colab_type": "text"
      },
      "source": [
        "# Try to test on real data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "W68zMEFKtI2n",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!gdown https://drive.google.com/uc?id=1pyq-4MRnSMrtb9FxL-pzuzACQhm28erC"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nTdrzlsZ2h8q",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!gdown https://drive.google.com/uc?id=1Ptnq7oNlGFbghEjYFKml0z4ae_T44Mo6"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bLGdRzEdtNjV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "'load data'\n",
        "# data, samplerate = sf.read('bgm_free_audio.wav')\n",
        "data = load_data_array_from_npy('bgm_free_audio.npy')\n",
        "# data = load_data_array_from_npy(f\"gdrive/My Drive/VTB_model/new_audio_apex.npy\")\n",
        "'Here is the pipline'\n",
        "skip_length = 8000 * 94\n",
        "data = data[skip_length:]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-oa4j8sziwLr",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def data_normalization(data, ratio = 1.1):\n",
        "  norm = np.max(np.abs(data)) * ratio\n",
        "  # print(norm)\n",
        "  # pdb.set_trace()\n",
        "  data /= norm\n",
        "  return data"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yFpEvCQYtd3A",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "samplerate = 8000 * 3\n",
        "total_second_number = int(len(data)//samplerate)\n",
        "'iterate over the data'\n",
        "splitted_res = []\n",
        "for i in range(20):\n",
        "  start_point = i * samplerate\n",
        "  end_point = (i + 1) * samplerate\n",
        "  strided_data = data[start_point:end_point]\n",
        "  assert len(strided_data) == samplerate\n",
        "  \n",
        "  strided_data = data_normalization(strided_data, ratio=1.1)\n",
        "  mixed_mag, mixed_phase = audio.wav2spec(strided_data)\n",
        "  \n",
        "  label_0_res = model.predict([np.expand_dims(mixed_mag, axis=0), np.expand_dims(label_0_embeddings, axis=0)])\n",
        "  label_1_res = model.predict([np.expand_dims(mixed_mag, axis=0), np.expand_dims(label_1_embeddings, axis=0)])\n",
        "\n",
        "  reconstruct_wav(mixed_mag, mixed_phase, \"Mixed voice\")\n",
        "  reconstruct_wav(label_0_res[0], mixed_phase, f\"prediction: label 0\")\n",
        "  reconstruct_wav(label_1_res[0], mixed_phase, f\"prediction: label 1\")\n",
        "  # 'normalize the data'\n",
        "  # if i > 20:\n",
        "  #   break"
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}