tacotron2_tts.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "tacotron2_tts.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyNHAzoLfsSJ5LQnX+8E741e",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/pavs315/TTS/blob/main/tacotron2_tts.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "IwGl8KSRB2NP",
        "outputId": "f9d7718c-82c1-4804-c472-dcecd00a2a98"
      },
      "source": [
        "!apt-get install -y libsndfile1\n",
        "!pip install unidecode"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Reading package lists... Done\n",
            "Building dependency tree       \n",
            "Reading state information... Done\n",
            "libsndfile1 is already the newest version (1.0.28-4ubuntu0.18.04.1).\n",
            "0 upgraded, 0 newly installed, 0 to remove and 40 not upgraded.\n",
            "Collecting unidecode\n",
            "  Downloading Unidecode-1.2.0-py2.py3-none-any.whl (241 kB)\n",
            "\u001b[K     |████████████████████████████████| 241 kB 8.2 MB/s \n",
            "\u001b[?25hInstalling collected packages: unidecode\n",
            "Successfully installed unidecode-1.2.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xdqt49-OCVEQ",
        "outputId": "c24026da-1223-424f-fab9-9d441d63a9a0"
      },
      "source": [
        "import torch\n",
        "tacotron2 = torch.hub.load('NVIDIA/DeepLearningExamples:torchhub', 'nvidia_tacotron2', model_math='fp16')\n",
        "tacotron2 = tacotron2.to('cuda')\n",
        "tacotron2.eval()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading: \"https://github.com/NVIDIA/DeepLearningExamples/archive/torchhub.zip\" to /root/.cache/torch/hub/torchhub.zip\n",
            "Downloading checkpoint from https://api.ngc.nvidia.com/v2/models/nvidia/tacotron2_pyt_ckpt_amp/versions/19.09.0/files/nvidia_tacotron2pyt_fp16_20190427\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Tacotron2(\n",
              "  (embedding): Embedding(148, 512)\n",
              "  (encoder): Encoder(\n",
              "    (convolutions): ModuleList(\n",
              "      (0): Sequential(\n",
              "        (0): ConvNorm(\n",
              "          (conv): Conv1d(512, 512, kernel_size=(5,), stride=(1,), padding=(2,))\n",
              "        )\n",
              "        (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
              "      )\n",
              "      (1): Sequential(\n",
              "        (0): ConvNorm(\n",
              "          (conv): Conv1d(512, 512, kernel_size=(5,), stride=(1,), padding=(2,))\n",
              "        )\n",
              "        (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
              "      )\n",
              "      (2): Sequential(\n",
              "        (0): ConvNorm(\n",
              "          (conv): Conv1d(512, 512, kernel_size=(5,), stride=(1,), padding=(2,))\n",
              "        )\n",
              "        (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
              "      )\n",
              "    )\n",
              "    (lstm): LSTM(512, 256, batch_first=True, bidirectional=True)\n",
              "  )\n",
              "  (decoder): Decoder(\n",
              "    (prenet): Prenet(\n",
              "      (layers): ModuleList(\n",
              "        (0): LinearNorm(\n",
              "          (linear_layer): Linear(in_features=80, out_features=256, bias=False)\n",
              "        )\n",
              "        (1): LinearNorm(\n",
              "          (linear_layer): Linear(in_features=256, out_features=256, bias=False)\n",
              "        )\n",
              "      )\n",
              "    )\n",
              "    (attention_rnn): LSTMCell(768, 1024)\n",
              "    (attention_layer): Attention(\n",
              "      (query_layer): LinearNorm(\n",
              "        (linear_layer): Linear(in_features=1024, out_features=128, bias=False)\n",
              "      )\n",
              "      (memory_layer): LinearNorm(\n",
              "        (linear_layer): Linear(in_features=512, out_features=128, bias=False)\n",
              "      )\n",
              "      (v): LinearNorm(\n",
              "        (linear_layer): Linear(in_features=128, out_features=1, bias=False)\n",
              "      )\n",
              "      (location_layer): LocationLayer(\n",
              "        (location_conv): ConvNorm(\n",
              "          (conv): Conv1d(2, 32, kernel_size=(31,), stride=(1,), padding=(15,), bias=False)\n",
              "        )\n",
              "        (location_dense): LinearNorm(\n",
              "          (linear_layer): Linear(in_features=32, out_features=128, bias=False)\n",
              "        )\n",
              "      )\n",
              "    )\n",
              "    (decoder_rnn): LSTMCell(1536, 1024, bias=1)\n",
              "    (linear_projection): LinearNorm(\n",
              "      (linear_layer): Linear(in_features=1536, out_features=80, bias=True)\n",
              "    )\n",
              "    (gate_layer): LinearNorm(\n",
              "      (linear_layer): Linear(in_features=1536, out_features=1, bias=True)\n",
              "    )\n",
              "  )\n",
              "  (postnet): Postnet(\n",
              "    (convolutions): ModuleList(\n",
              "      (0): Sequential(\n",
              "        (0): ConvNorm(\n",
              "          (conv): Conv1d(80, 512, kernel_size=(5,), stride=(1,), padding=(2,))\n",
              "        )\n",
              "        (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
              "      )\n",
              "      (1): Sequential(\n",
              "        (0): ConvNorm(\n",
              "          (conv): Conv1d(512, 512, kernel_size=(5,), stride=(1,), padding=(2,))\n",
              "        )\n",
              "        (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
              "      )\n",
              "      (2): Sequential(\n",
              "        (0): ConvNorm(\n",
              "          (conv): Conv1d(512, 512, kernel_size=(5,), stride=(1,), padding=(2,))\n",
              "        )\n",
              "        (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
              "      )\n",
              "      (3): Sequential(\n",
              "        (0): ConvNorm(\n",
              "          (conv): Conv1d(512, 512, kernel_size=(5,), stride=(1,), padding=(2,))\n",
              "        )\n",
              "        (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
              "      )\n",
              "      (4): Sequential(\n",
              "        (0): ConvNorm(\n",
              "          (conv): Conv1d(512, 80, kernel_size=(5,), stride=(1,), padding=(2,))\n",
              "        )\n",
              "        (1): BatchNorm1d(80, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
              "      )\n",
              "    )\n",
              "  )\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9rq2j2DQCuBK",
        "outputId": "a5ab1a4b-5c96-4954-cf77-c193b2f0e32a"
      },
      "source": [
        "waveglow = torch.hub.load('NVIDIA/DeepLearningExamples:torchhub', 'nvidia_waveglow', model_math='fp16')\n",
        "waveglow = waveglow.remove_weightnorm(waveglow)\n",
        "waveglow = waveglow.to('cuda')\n",
        "waveglow.eval()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using cache found in /root/.cache/torch/hub/NVIDIA_DeepLearningExamples_torchhub\n",
            "Downloading checkpoint from https://api.ngc.nvidia.com/v2/models/nvidia/waveglow_ckpt_amp/versions/19.09.0/files/nvidia_waveglowpyt_fp16_20190427\n",
            "/root/.cache/torch/hub/NVIDIA_DeepLearningExamples_torchhub/PyTorch/SpeechSynthesis/Tacotron2/waveglow/model.py:55: UserWarning: torch.qr is deprecated in favor of torch.linalg.qr and will be removed in a future PyTorch release.\n",
            "The boolean parameter 'some' has been replaced with a string parameter 'mode'.\n",
            "Q, R = torch.qr(A, some)\n",
            "should be replaced with\n",
            "Q, R = torch.linalg.qr(A, 'reduced' if some else 'complete') (Triggered internally at  /pytorch/aten/src/ATen/native/BatchLinearAlgebra.cpp:1940.)\n",
            "  W = torch.qr(torch.FloatTensor(c, c).normal_())[0]\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "WaveGlow(\n",
              "  (upsample): ConvTranspose1d(80, 80, kernel_size=(1024,), stride=(256,))\n",
              "  (WN): ModuleList(\n",
              "    (0): WN(\n",
              "      (in_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(2,), dilation=(2,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(4,), dilation=(4,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(8,), dilation=(8,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(16,), dilation=(16,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(32,), dilation=(32,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(64,), dilation=(64,))\n",
              "        (7): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(128,), dilation=(128,))\n",
              "      )\n",
              "      (res_skip_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(512, 512, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (cond_layers): ModuleList(\n",
              "        (0): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (start): Conv1d(4, 512, kernel_size=(1,), stride=(1,))\n",
              "      (end): Conv1d(512, 8, kernel_size=(1,), stride=(1,))\n",
              "    )\n",
              "    (1): WN(\n",
              "      (in_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(2,), dilation=(2,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(4,), dilation=(4,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(8,), dilation=(8,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(16,), dilation=(16,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(32,), dilation=(32,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(64,), dilation=(64,))\n",
              "        (7): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(128,), dilation=(128,))\n",
              "      )\n",
              "      (res_skip_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(512, 512, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (cond_layers): ModuleList(\n",
              "        (0): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (start): Conv1d(4, 512, kernel_size=(1,), stride=(1,))\n",
              "      (end): Conv1d(512, 8, kernel_size=(1,), stride=(1,))\n",
              "    )\n",
              "    (2): WN(\n",
              "      (in_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(2,), dilation=(2,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(4,), dilation=(4,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(8,), dilation=(8,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(16,), dilation=(16,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(32,), dilation=(32,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(64,), dilation=(64,))\n",
              "        (7): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(128,), dilation=(128,))\n",
              "      )\n",
              "      (res_skip_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(512, 512, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (cond_layers): ModuleList(\n",
              "        (0): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (start): Conv1d(4, 512, kernel_size=(1,), stride=(1,))\n",
              "      (end): Conv1d(512, 8, kernel_size=(1,), stride=(1,))\n",
              "    )\n",
              "    (3): WN(\n",
              "      (in_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(2,), dilation=(2,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(4,), dilation=(4,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(8,), dilation=(8,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(16,), dilation=(16,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(32,), dilation=(32,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(64,), dilation=(64,))\n",
              "        (7): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(128,), dilation=(128,))\n",
              "      )\n",
              "      (res_skip_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(512, 512, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (cond_layers): ModuleList(\n",
              "        (0): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (start): Conv1d(4, 512, kernel_size=(1,), stride=(1,))\n",
              "      (end): Conv1d(512, 8, kernel_size=(1,), stride=(1,))\n",
              "    )\n",
              "    (4): WN(\n",
              "      (in_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(2,), dilation=(2,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(4,), dilation=(4,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(8,), dilation=(8,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(16,), dilation=(16,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(32,), dilation=(32,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(64,), dilation=(64,))\n",
              "        (7): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(128,), dilation=(128,))\n",
              "      )\n",
              "      (res_skip_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(512, 512, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (cond_layers): ModuleList(\n",
              "        (0): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (start): Conv1d(3, 512, kernel_size=(1,), stride=(1,))\n",
              "      (end): Conv1d(512, 6, kernel_size=(1,), stride=(1,))\n",
              "    )\n",
              "    (5): WN(\n",
              "      (in_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(2,), dilation=(2,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(4,), dilation=(4,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(8,), dilation=(8,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(16,), dilation=(16,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(32,), dilation=(32,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(64,), dilation=(64,))\n",
              "        (7): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(128,), dilation=(128,))\n",
              "      )\n",
              "      (res_skip_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(512, 512, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (cond_layers): ModuleList(\n",
              "        (0): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (start): Conv1d(3, 512, kernel_size=(1,), stride=(1,))\n",
              "      (end): Conv1d(512, 6, kernel_size=(1,), stride=(1,))\n",
              "    )\n",
              "    (6): WN(\n",
              "      (in_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(2,), dilation=(2,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(4,), dilation=(4,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(8,), dilation=(8,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(16,), dilation=(16,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(32,), dilation=(32,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(64,), dilation=(64,))\n",
              "        (7): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(128,), dilation=(128,))\n",
              "      )\n",
              "      (res_skip_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(512, 512, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (cond_layers): ModuleList(\n",
              "        (0): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (start): Conv1d(3, 512, kernel_size=(1,), stride=(1,))\n",
              "      (end): Conv1d(512, 6, kernel_size=(1,), stride=(1,))\n",
              "    )\n",
              "    (7): WN(\n",
              "      (in_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(2,), dilation=(2,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(4,), dilation=(4,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(8,), dilation=(8,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(16,), dilation=(16,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(32,), dilation=(32,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(64,), dilation=(64,))\n",
              "        (7): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(128,), dilation=(128,))\n",
              "      )\n",
              "      (res_skip_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(512, 512, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (cond_layers): ModuleList(\n",
              "        (0): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (start): Conv1d(3, 512, kernel_size=(1,), stride=(1,))\n",
              "      (end): Conv1d(512, 6, kernel_size=(1,), stride=(1,))\n",
              "    )\n",
              "    (8): WN(\n",
              "      (in_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(2,), dilation=(2,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(4,), dilation=(4,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(8,), dilation=(8,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(16,), dilation=(16,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(32,), dilation=(32,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(64,), dilation=(64,))\n",
              "        (7): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(128,), dilation=(128,))\n",
              "      )\n",
              "      (res_skip_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(512, 512, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (cond_layers): ModuleList(\n",
              "        (0): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (start): Conv1d(2, 512, kernel_size=(1,), stride=(1,))\n",
              "      (end): Conv1d(512, 4, kernel_size=(1,), stride=(1,))\n",
              "    )\n",
              "    (9): WN(\n",
              "      (in_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(2,), dilation=(2,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(4,), dilation=(4,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(8,), dilation=(8,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(16,), dilation=(16,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(32,), dilation=(32,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(64,), dilation=(64,))\n",
              "        (7): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(128,), dilation=(128,))\n",
              "      )\n",
              "      (res_skip_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(512, 512, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (cond_layers): ModuleList(\n",
              "        (0): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (start): Conv1d(2, 512, kernel_size=(1,), stride=(1,))\n",
              "      (end): Conv1d(512, 4, kernel_size=(1,), stride=(1,))\n",
              "    )\n",
              "    (10): WN(\n",
              "      (in_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(2,), dilation=(2,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(4,), dilation=(4,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(8,), dilation=(8,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(16,), dilation=(16,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(32,), dilation=(32,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(64,), dilation=(64,))\n",
              "        (7): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(128,), dilation=(128,))\n",
              "      )\n",
              "      (res_skip_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(512, 512, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (cond_layers): ModuleList(\n",
              "        (0): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (start): Conv1d(2, 512, kernel_size=(1,), stride=(1,))\n",
              "      (end): Conv1d(512, 4, kernel_size=(1,), stride=(1,))\n",
              "    )\n",
              "    (11): WN(\n",
              "      (in_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(2,), dilation=(2,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(4,), dilation=(4,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(8,), dilation=(8,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(16,), dilation=(16,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(32,), dilation=(32,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(64,), dilation=(64,))\n",
              "        (7): Conv1d(512, 1024, kernel_size=(3,), stride=(1,), padding=(128,), dilation=(128,))\n",
              "      )\n",
              "      (res_skip_layers): ModuleList(\n",
              "        (0): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(512, 512, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (cond_layers): ModuleList(\n",
              "        (0): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (1): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (2): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (3): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (4): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (5): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (6): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "        (7): Conv1d(640, 1024, kernel_size=(1,), stride=(1,))\n",
              "      )\n",
              "      (start): Conv1d(2, 512, kernel_size=(1,), stride=(1,))\n",
              "      (end): Conv1d(512, 4, kernel_size=(1,), stride=(1,))\n",
              "    )\n",
              "  )\n",
              "  (convinv): ModuleList(\n",
              "    (0): Invertible1x1Conv(\n",
              "      (conv): Conv1d(8, 8, kernel_size=(1,), stride=(1,), bias=False)\n",
              "    )\n",
              "    (1): Invertible1x1Conv(\n",
              "      (conv): Conv1d(8, 8, kernel_size=(1,), stride=(1,), bias=False)\n",
              "    )\n",
              "    (2): Invertible1x1Conv(\n",
              "      (conv): Conv1d(8, 8, kernel_size=(1,), stride=(1,), bias=False)\n",
              "    )\n",
              "    (3): Invertible1x1Conv(\n",
              "      (conv): Conv1d(8, 8, kernel_size=(1,), stride=(1,), bias=False)\n",
              "    )\n",
              "    (4): Invertible1x1Conv(\n",
              "      (conv): Conv1d(6, 6, kernel_size=(1,), stride=(1,), bias=False)\n",
              "    )\n",
              "    (5): Invertible1x1Conv(\n",
              "      (conv): Conv1d(6, 6, kernel_size=(1,), stride=(1,), bias=False)\n",
              "    )\n",
              "    (6): Invertible1x1Conv(\n",
              "      (conv): Conv1d(6, 6, kernel_size=(1,), stride=(1,), bias=False)\n",
              "    )\n",
              "    (7): Invertible1x1Conv(\n",
              "      (conv): Conv1d(6, 6, kernel_size=(1,), stride=(1,), bias=False)\n",
              "    )\n",
              "    (8): Invertible1x1Conv(\n",
              "      (conv): Conv1d(4, 4, kernel_size=(1,), stride=(1,), bias=False)\n",
              "    )\n",
              "    (9): Invertible1x1Conv(\n",
              "      (conv): Conv1d(4, 4, kernel_size=(1,), stride=(1,), bias=False)\n",
              "    )\n",
              "    (10): Invertible1x1Conv(\n",
              "      (conv): Conv1d(4, 4, kernel_size=(1,), stride=(1,), bias=False)\n",
              "    )\n",
              "    (11): Invertible1x1Conv(\n",
              "      (conv): Conv1d(4, 4, kernel_size=(1,), stride=(1,), bias=False)\n",
              "    )\n",
              "  )\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "D1kYimYECyKr",
        "outputId": "2aa05607-282b-436c-ed78-74bc947ea952"
      },
      "source": [
        "import unidecode\n",
        "text = \"is this thing working\"\n",
        "utils = torch.hub.load('NVIDIA/DeepLearningExamples:torchhub', 'nvidia_tts_utils')\n",
        "sequences, lengths = utils.prepare_input_sequence([text])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using cache found in /root/.cache/torch/hub/NVIDIA_DeepLearningExamples_torchhub\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GVO_s5tdDBT3"
      },
      "source": [
        "with torch.no_grad():\n",
        "    mel, _, _ = tacotron2.infer(sequences, lengths)\n",
        "    audio = waveglow.infer(mel)\n",
        "audio_numpy = audio[0].data.cpu().numpy()\n",
        "rate = 22050"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 76
        },
        "id": "yyLDvtmhDM-B",
        "outputId": "57daac25-852a-487a-ede1-e1d8c8a5995e"
      },
      "source": [
        "import IPython\n",
        "from IPython.display import Audio\n",
        "Audio(audio_numpy, rate=rate)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "                <audio controls=\"controls\" >\n",
              "                    <source src=\"data:audio/wav;base64,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\" type=\"audio/wav\" />\n",
              "                    Your browser does not support the audio element.\n",
              "                </audio>\n",
              "              "
            ],
            "text/plain": [
              "<IPython.lib.display.Audio object>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    }
  ]
}