bkp_FakeYou_Tacotron_2_(w_ARPAbet)_Training.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ODI651CWUdFs"
      },
      "source": [
        "# <b><font color=\"pink\" size=\"+2\"> **FakeYou Tacotron 2 *(w/ARPAbet)* Training Colab**\n",
        "\n",
        "### [Tacotron 2](https://github.com/NVIDIA/tacotron2)\n",
        "![FakeYou_Logo.png](data:image/png;base64,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)\n",
        "### [FakeYou](https://fakeyou.com/)\n",
        "---\n",
        "\n",
        "### **Disclaimer:**  You may encounter difficulties when uploading files (> 1MB) if you use firefox.\n",
        "---\n",
        "\n",
        "#### Special thanks to Cookie from the Pony Preservation Project\n",
        "###### Slight modifications by mega b, anonymous and justinjohn-03"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iQBUZTuMogui"
      },
      "source": [
        "# <b><font color=\"pink\" size=\"+2\"> Optional cells (unhide by clicking the arrow to the left)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "cellView": "form",
        "id": "75SFQOP_Wnpz"
      },
      "outputs": [],
      "source": [
        "#@markdown # Check GPU type\n",
        "#@markdown ### Factory reset runtime if you don't have the desired GPU.\n",
        "\n",
        "#@markdown ---\n",
        "\n",
        "#@markdown ## It is recommended to not use the K80\n",
        "\n",
        "!nvidia-smi -L\n",
        "\n",
        "#@markdown ---"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "cellView": "form",
        "id": "VKYvFCrUhI4O"
      },
      "outputs": [],
      "source": [
        "#@markdown # Anti-Disconnect for Google Colab\n",
        "#@markdown ## Run this to stop it from disconnecting automatically (will disconnect after 6 - 12 hours, though.)\n",
        "\n",
        "import IPython\n",
        "js_code = '''\n",
        "function ClickConnect(){\n",
        "console.log(\"Working\");\n",
        "document.querySelector(\"colab-toolbar-button#connect\").click()\n",
        "}\n",
        "setInterval(ClickConnect,60000)\n",
        "'''\n",
        "display(IPython.display.Javascript(js_code))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "cellView": "form",
        "id": "8LavWMVp3fMt"
      },
      "outputs": [],
      "source": [
        "#@markdown # Unzip wavs from Gdrive\n",
        "#@markdown ###**(run this cell after running step 3)**\n",
        "#@markdown ---\n",
        "#@markdown ### If you have a lot of wav files, then zip them all into one file localy on your system, then upload it and copy the path.\n",
        "#@markdown ---\n",
        "\n",
        "import os\n",
        "\n",
        "zip_file_path = \"\\\"/content/drive/MyDrive/Vocoderstuff/JohnnyDepp.zip\\\"\" #@param {type:\"string\"}\n",
        "\n",
        "if os.path.isdir(\"/content/tacotron2/wavs\"):\n",
        "  !unzip $zip_file_path -d /content/tacotron2/wavs\n",
        "else:\n",
        "  print(\"Failed: You must run the second step in the training first, before running this.\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OhZSeuWHoj6g"
      },
      "source": [
        "# **Setup**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "cellView": "form",
        "id": "z6aq6vRzqEV9"
      },
      "outputs": [],
      "source": [
        "#@markdown ## <b><font color=\"pink\" size=\"+2\"> Step:1 **Mount your Google Drive**\n",
        "\n",
        "from google.colab import drive\n",
        "drive.mount('drive')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "cellView": "form",
        "id": "7TPdPew-YZVA"
      },
      "outputs": [],
      "source": [
        "#@markdown ## <b><font color=\"pink\" size=\"+2\"> **Step:2** **Install Tacotron2 (w/ARPAbet)**\n",
        "import os\n",
        "%cd /content/\n",
        "!pip install tensorflow==1.15.2\n",
        "if not os.path.isdir(\"/content/tacotron2/\"):\n",
        "  print(\"Cloning NVIDIA/tacotron2\")\n",
        "  !git clone https://github.com/NVIDIA/tacotron2.git\n",
        "  %cd /content/tacotron2/\n",
        "  !git submodule init\n",
        "  !git submodule update\n",
        "%cd /content/tacotron2/\n",
        "#NVIDIA's requirements\n",
        "#I believe Colab gives us PyTorch and TF by default so we don't need anything else\n",
        "#Versions specified in requirements.txt have conflicts so that's why we simply get current versions \n",
        "print(\"Downloading tacotron2 requirements\")\n",
        "!pip install matplotlib numpy inflect librosa scipy Unidecode pillow\n",
        "#Our requirements\n",
        "#We'll need gdown to download some really cool things\n",
        "!pip install gdown\n",
        "import gdown\n",
        "!git submodule init\n",
        "!git submodule update\n",
        "!pip install -q unidecode tensorboardX\n",
        "!apt-get install pv\n",
        "!apt-get install jq\n",
        "!wget https://raw.githubusercontent.com/tonikelope/megadown/master/megadown -O megadown.sh\n",
        "!chmod 755 megadown.sh\n",
        "#Download NVIDIA's LJSpeech model\n",
        "!wget \"https://github.com/justinjohn0306/tacotron2/releases/download/model/tacotron2_statedict.pt\" -O pretrained_model\n",
        "if not os.path.isfile(\"/content/tacotron2/text/merged.dict.txt\"):\n",
        "  print(\"Applying cmudict patch\")\n",
        "  !curl https://cdn.discordapp.com/attachments/820353681567907901/865742324084244480/tacotron2-cmudict-patch.zip -o /content/tacotron2-cmudict-patch.zip\n",
        "  !unzip -o /content/tacotron2-cmudict-patch.zip -d /content/tacotron2/text/\n",
        "  !mv /content/tacotron2/text/merged.dict.txt /content/tacotron2/"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "cellView": "form",
        "id": "TFVNCB2BYh1Y"
      },
      "outputs": [],
      "source": [
        "#@markdown ## <b><font color=\"pink\" size=\"+2\"> Step:3 **Apply additional ARPAbet patches**\n",
        "patches = [\n",
        "  (\"/content/tacotron2/train.patch\", \"\"\"17a18\n",
        "> from tqdm import tqdm\n",
        "18a20,26\n",
        "> from plotting_utils import plot_alignment_to_numpy\n",
        "> from IPython import display\n",
        "> from PIL import Image\n",
        "> \n",
        "> def plot_something_to_ipython(alignment):\n",
        ">   numpoop = plot_alignment_to_numpy(alignment)\n",
        ">   display.display(Image.fromarray(numpoop))\n",
        "28,29c36,37\n",
        "<     assert torch.cuda.is_available(), \"Distributed mode requires CUDA.\"\n",
        "<     print(\"Initializing Distributed\")\n",
        "---\n",
        ">     assert torch.cuda.is_available(), \"distributed mode requires CUDA\"\n",
        ">     print(\"initializing distributed\")\n",
        "39c47\n",
        "<     print(\"Done initializing distributed\")\n",
        "---\n",
        ">     print(\"done initializing distributed\")\n",
        "86c94\n",
        "<     print(\"Warm starting model from checkpoint '{}'\".format(checkpoint_path))\n",
        "---\n",
        ">     print(\"warm starting model from checkpoint '{}'\".format(checkpoint_path))\n",
        "101c109\n",
        "<     print(\"Loading checkpoint '{}'\".format(checkpoint_path))\n",
        "---\n",
        ">     print(\"loading checkpoint '{}'\".format(checkpoint_path))\n",
        "107c115\n",
        "<     print(\"Loaded checkpoint '{}' from iteration {}\" .format(\n",
        "---\n",
        ">     print(\"loaded checkpoint '{}' from iteration {}\" .format(\n",
        "113c121\n",
        "<     print(\"Saving model and optimizer state at iteration {} to {}\".format(\n",
        "---\n",
        ">     print(\"saving model and optimizer state at iteration {} to {}\".format(\n",
        "145c153\n",
        "<         print(\"Validation loss {}: {:9f}  \".format(iteration, val_loss))\n",
        "---\n",
        ">         print(\"validation loss {}: {:9f}  \".format(iteration, val_loss))\n",
        "146a155,161\n",
        ">         \n",
        ">         # plot_something_to_ipython\n",
        ">         # https://github.com/NVIDIA/tacotron2/blob/master/logger.py\n",
        ">         _, _, _, alignments = y_pred\n",
        ">         from random import randint\n",
        ">         idx = randint(0, alignments.size(0)-1)\n",
        ">         plot_something_to_ipython(alignments[idx].data.cpu().numpy().T)\n",
        "169a185\n",
        ">     print(\"starting with {} learning rate\".format(learning_rate))\n",
        "207,208c223,225\n",
        "<         print(\"Epoch: {}\".format(epoch))\n",
        "<         for i, batch in enumerate(train_loader):\n",
        "---\n",
        ">         print(\"starting epoch {} at iteration {}\".format(epoch, iteration))\n",
        ">         epochstart = time.perf_counter()\n",
        ">         for i, batch in tqdm(enumerate(train_loader)):\n",
        "240,241c257,258\n",
        "<                 print(\"Train loss {} {:.6f} Grad Norm {:.6f} {:.2f}s/it\".format(\n",
        "<                     iteration, reduced_loss, grad_norm, duration))\n",
        "---\n",
        ">                 #print(\"Train loss {} {:.6f} Grad Norm {:.6f} {:.2f}s/it\".format(\n",
        ">                 #    iteration, reduced_loss, grad_norm, duration))\n",
        "243a261\n",
        ">             from random import random\n",
        "245,246c263,266\n",
        "<             if not is_overflow and (iteration % hparams.iters_per_checkpoint == 0):\n",
        "<                 validate(model, criterion, valset, iteration,\n",
        "---\n",
        ">             iteration += 1\n",
        ">         print(\"\\nepoch {} took {} seconds\".format(epoch, time.perf_counter() - epochstart))\n",
        ">         #what could possibly go wrong\n",
        ">         validate(model, criterion, valset, iteration,\n",
        "248a269\n",
        ">         if not is_overflow and (random() > 0.66): #(iteration % hparams.iters_per_checkpoint == 0):\n",
        "251,256c272,280\n",
        "<                         output_directory, \"checkpoint_{}\".format(iteration))\n",
        "<                     save_checkpoint(model, optimizer, learning_rate, iteration,\n",
        "<                                     checkpoint_path)\n",
        "< \n",
        "<             iteration += 1\n",
        "< \n",
        "---\n",
        ">                         output_directory, hparams.model_name)\n",
        ">                     try:\n",
        ">                         save_checkpoint(model, optimizer, learning_rate, iteration,\n",
        ">                                         checkpoint_path)\n",
        ">                     except KeyboardInterrupt:\n",
        ">                         print(\"you probably shouldnt ^C while im saving\")\n",
        ">                         save_checkpoint(model, optimizer, learning_rate, iteration,\n",
        ">                                         checkpoint_path)\n",
        ">                         print(\"ok it should be fine now\")\n",
        "\n",
        "\"\"\"),\n",
        "  (\"/content/tacotron2/plotting_utils.patch\", \"\"\"5c5\n",
        "< \n",
        "---\n",
        "> import io\n",
        "9,10c9,13\n",
        "<     data = np.fromstring(fig.canvas.tostring_rgb(), dtype=np.uint8, sep='')\n",
        "<     data = data.reshape(fig.canvas.get_width_height()[::-1] + (3,))\n",
        "---\n",
        ">     # no obvious way to make it rgba... https://github.com/matplotlib/matplotlib/issues/5336#issuecomment-388736185\n",
        ">     buf = io.BytesIO()\n",
        ">     fig.savefig(buf, format=\"rgba\")\n",
        ">     data = np.fromstring(buf.getvalue(), dtype=np.uint8, sep='')\n",
        ">     data = data.reshape(fig.canvas.get_width_height()[::-1] + (4,))\n",
        "15c18\n",
        "<     fig, ax = plt.subplots(figsize=(6, 4))\n",
        "---\n",
        ">     fig, ax = plt.subplots(figsize=(9, 6))\n",
        "17c20\n",
        "<                    interpolation='none')\n",
        "---\n",
        ">                    interpolation='none', cmap='inferno')\n",
        "35c38\n",
        "<                    interpolation='none')\n",
        "---\n",
        ">                    interpolation='none', cmap='inferno')\n",
        "\n",
        "\"\"\")\n",
        "]\n",
        "for i, v in enumerate(patches):\n",
        "  to = v[0]\n",
        "  co = v[1]\n",
        "  with open(to, \"w\") as file:\n",
        "    file.write(co)\n",
        "\n",
        "from glob import glob \n",
        "for x in glob(\"*.patch\"):\n",
        "  base = x[:-6]\n",
        "  patch = x\n",
        "  py = base+\".py\"\n",
        "  !patch {py} {patch}\n",
        "  import time\n",
        "import argparse\n",
        "import math\n",
        "from numpy import finfo\n",
        "\n",
        "import torch\n",
        "from distributed import apply_gradient_allreduce\n",
        "import torch.distributed as dist\n",
        "from torch.utils.data.distributed import DistributedSampler\n",
        "from torch.utils.data import DataLoader\n",
        "\n",
        "from model import Tacotron2\n",
        "from data_utils import TextMelLoader, TextMelCollate\n",
        "from loss_function import Tacotron2Loss\n",
        "from logger import Tacotron2Logger\n",
        "from hparams import create_hparams\n",
        " \n",
        "import random\n",
        "import numpy as np\n",
        "\n",
        "import layers\n",
        "from utils import load_wav_to_torch, load_filepaths_and_text\n",
        "from text import text_to_sequence\n",
        "from math import e\n",
        "#from tqdm import tqdm # Terminal\n",
        "#from tqdm import tqdm_notebook as tqdm # Legacy Notebook TQDM\n",
        "from tqdm.notebook import tqdm # Modern Notebook TQDM\n",
        "from distutils.dir_util import copy_tree\n",
        "import matplotlib.pylab as plt\n",
        "\n",
        "def download_from_google_drive(file_id, file_name):\n",
        "  # download a file from the Google Drive link\n",
        "  !rm -f ./cookie\n",
        "  !curl -c ./cookie -s -L \"https://drive.google.com/uc?export=download&id={file_id}\" > /dev/null\n",
        "  confirm_text = !awk '/download/ {print $NF}' ./cookie\n",
        "  confirm_text = confirm_text[0]\n",
        "  !curl -Lb ./cookie \"https://drive.google.com/uc?export=download&confirm={confirm_text}&id={file_id}\" -o {file_name}\n",
        "\n",
        "def create_mels():\n",
        "    print(\"Generating Mels\")\n",
        "    stft = layers.TacotronSTFT(\n",
        "                hparams.filter_length, hparams.hop_length, hparams.win_length,\n",
        "                hparams.n_mel_channels, hparams.sampling_rate, hparams.mel_fmin,\n",
        "                hparams.mel_fmax)\n",
        "    def save_mel(filename):\n",
        "        audio, sampling_rate = load_wav_to_torch(filename)\n",
        "        if sampling_rate != stft.sampling_rate:\n",
        "            raise ValueError(\"{} {} SR doesn't match target {} SR\".format(filename, \n",
        "                sampling_rate, stft.sampling_rate))\n",
        "        audio_norm = audio / hparams.max_wav_value\n",
        "        audio_norm = audio_norm.unsqueeze(0)\n",
        "        audio_norm = torch.autograd.Variable(audio_norm, requires_grad=False)\n",
        "        melspec = stft.mel_spectrogram(audio_norm)\n",
        "        melspec = torch.squeeze(melspec, 0).cpu().numpy()\n",
        "        np.save(filename.replace('.wav', ''), melspec)\n",
        "\n",
        "    import glob\n",
        "    wavs = glob.glob('wavs/*.wav')\n",
        "    for i in tqdm(wavs):\n",
        "        save_mel(i)\n",
        "\n",
        "\n",
        "def reduce_tensor(tensor, n_gpus):\n",
        "    rt = tensor.clone()\n",
        "    dist.all_reduce(rt, op=dist.reduce_op.SUM)\n",
        "    rt /= n_gpus\n",
        "    return rt\n",
        "\n",
        "\n",
        "def init_distributed(hparams, n_gpus, rank, group_name):\n",
        "    assert torch.cuda.is_available(), \"Distributed mode requires CUDA.\"\n",
        "    print(\"Initializing Distributed\")\n",
        "\n",
        "    # Set cuda device so everything is done on the right GPU.\n",
        "    torch.cuda.set_device(rank % torch.cuda.device_count())\n",
        "\n",
        "    # Initialize distributed communication\n",
        "    dist.init_process_group(\n",
        "        backend=hparams.dist_backend, init_method=hparams.dist_url,\n",
        "        world_size=n_gpus, rank=rank, group_name=group_name)\n",
        "\n",
        "    print(\"Done initializing distributed\")\n",
        "\n",
        "\n",
        "def prepare_dataloaders(hparams):\n",
        "    # Get data, data loaders and collate function ready\n",
        "    trainset = TextMelLoader(hparams.training_files, hparams)\n",
        "    valset = TextMelLoader(hparams.validation_files, hparams)\n",
        "    collate_fn = TextMelCollate(hparams.n_frames_per_step)\n",
        "\n",
        "    if hparams.distributed_run:\n",
        "        train_sampler = DistributedSampler(trainset)\n",
        "        shuffle = False\n",
        "    else:\n",
        "        train_sampler = None\n",
        "        shuffle = True\n",
        "\n",
        "    train_loader = DataLoader(trainset, num_workers=1, shuffle=shuffle,\n",
        "                              sampler=train_sampler,\n",
        "                              batch_size=hparams.batch_size, pin_memory=False,\n",
        "                              drop_last=True, collate_fn=collate_fn)\n",
        "    return train_loader, valset, collate_fn\n",
        "\n",
        "\n",
        "def prepare_directories_and_logger(output_directory, log_directory, rank):\n",
        "    if rank == 0:\n",
        "        if not os.path.isdir(output_directory):\n",
        "            os.makedirs(output_directory)\n",
        "            os.chmod(output_directory, 0o775)\n",
        "        logger = Tacotron2Logger(os.path.join(output_directory, log_directory))\n",
        "    else:\n",
        "        logger = None\n",
        "    return logger\n",
        "\n",
        "\n",
        "def load_model(hparams):\n",
        "    model = Tacotron2(hparams).cuda()\n",
        "    if hparams.fp16_run:\n",
        "        model.decoder.attention_layer.score_mask_value = finfo('float16').min\n",
        "\n",
        "    if hparams.distributed_run:\n",
        "        model = apply_gradient_allreduce(model)\n",
        "\n",
        "    return model\n",
        "\n",
        "\n",
        "def warm_start_model(checkpoint_path, model, ignore_layers):\n",
        "    assert os.path.isfile(checkpoint_path)\n",
        "    print(\"Warm starting model from checkpoint '{}'\".format(checkpoint_path))\n",
        "    checkpoint_dict = torch.load(checkpoint_path, map_location='cpu')\n",
        "    model_dict = checkpoint_dict['state_dict']\n",
        "    if len(ignore_layers) > 0:\n",
        "        model_dict = {k: v for k, v in model_dict.items()\n",
        "                      if k not in ignore_layers}\n",
        "        dummy_dict = model.state_dict()\n",
        "        dummy_dict.update(model_dict)\n",
        "        model_dict = dummy_dict\n",
        "    model.load_state_dict(model_dict)\n",
        "    return model\n",
        "\n",
        "\n",
        "def load_checkpoint(checkpoint_path, model, optimizer):\n",
        "    assert os.path.isfile(checkpoint_path)\n",
        "    print(\"Loading checkpoint '{}'\".format(checkpoint_path))\n",
        "    checkpoint_dict = torch.load(checkpoint_path, map_location='cpu')\n",
        "    model.load_state_dict(checkpoint_dict['state_dict'])\n",
        "    optimizer.load_state_dict(checkpoint_dict['optimizer'])\n",
        "    learning_rate = checkpoint_dict['learning_rate']\n",
        "    iteration = checkpoint_dict['iteration']\n",
        "    print(\"Loaded checkpoint '{}' from iteration {}\" .format(\n",
        "        checkpoint_path, iteration))\n",
        "    return model, optimizer, learning_rate, iteration\n",
        "\n",
        "\n",
        "def save_checkpoint(model, optimizer, learning_rate, iteration, filepath):\n",
        "    import random\n",
        "    if True:\n",
        "        print(\"Saving model and optimizer state at iteration {} to {}\".format(\n",
        "            iteration, filepath))\n",
        "        try:\n",
        "            torch.save({'iteration': iteration,\n",
        "                    'state_dict': model.state_dict(),\n",
        "                    'optimizer': optimizer.state_dict(),\n",
        "                    'learning_rate': learning_rate}, filepath)\n",
        "        except KeyboardInterrupt:\n",
        "            print(\"interrupt received while saving, waiting for save to complete.\")\n",
        "            torch.save({'iteration': iteration,'state_dict': model.state_dict(),'optimizer': optimizer.state_dict(),'learning_rate': learning_rate}, filepath)\n",
        "        print(\"Model Saved\")\n",
        "\n",
        "def plot_alignment(alignment, info=None):\n",
        "    %matplotlib inline\n",
        "    fig, ax = plt.subplots(figsize=(int(alignment_graph_width/100), int(alignment_graph_height/100)))\n",
        "    im = ax.imshow(alignment, cmap='inferno', aspect='auto', origin='lower',\n",
        "                   interpolation='none')\n",
        "    ax.autoscale(enable=True, axis=\"y\", tight=True)\n",
        "    fig.colorbar(im, ax=ax)\n",
        "    xlabel = 'Decoder timestep'\n",
        "    if info is not None:\n",
        "        xlabel += '\\n\\n' + info\n",
        "    plt.xlabel(xlabel)\n",
        "    plt.ylabel('Encoder timestep')\n",
        "    plt.tight_layout()\n",
        "    fig.canvas.draw()\n",
        "    plt.show()\n",
        "\n",
        "def validate(model, criterion, valset, iteration, batch_size, n_gpus,\n",
        "             collate_fn, logger, distributed_run, rank, epoch, start_eposh, learning_rate):\n",
        "    \"\"\"Handles all the validation scoring and printing\"\"\"\n",
        "    model.eval()\n",
        "    with torch.no_grad():\n",
        "        val_sampler = DistributedSampler(valset) if distributed_run else None\n",
        "        val_loader = DataLoader(valset, sampler=val_sampler, num_workers=1,\n",
        "                                shuffle=False, batch_size=batch_size,\n",
        "                                pin_memory=False, collate_fn=collate_fn)\n",
        "\n",
        "        val_loss = 0.0\n",
        "        for i, batch in enumerate(val_loader):\n",
        "            x, y = model.parse_batch(batch)\n",
        "            y_pred = model(x)\n",
        "            loss = criterion(y_pred, y)\n",
        "            if distributed_run:\n",
        "                reduced_val_loss = reduce_tensor(loss.data, n_gpus).item()\n",
        "            else:\n",
        "                reduced_val_loss = loss.item()\n",
        "            val_loss += reduced_val_loss\n",
        "        val_loss = val_loss / (i + 1)\n",
        "\n",
        "    model.train()\n",
        "    if rank == 0:\n",
        "        print(\"Epoch: {} Validation loss {}: {:9f}  Time: {:.1f}m LR: {:.6f}\".format(epoch, iteration, val_loss,(time.perf_counter()-start_eposh)/60, learning_rate))\n",
        "        logger.log_validation(val_loss, model, y, y_pred, iteration)\n",
        "        if hparams.show_alignments:\n",
        "            %matplotlib inline\n",
        "            _, mel_outputs, gate_outputs, alignments = y_pred\n",
        "            idx = random.randint(0, alignments.size(0) - 1)\n",
        "            plot_alignment(alignments[idx].data.cpu().numpy().T)\n",
        "\n",
        "def train(output_directory, log_directory, checkpoint_path, warm_start, n_gpus,\n",
        "          rank, group_name, hparams, log_directory2):\n",
        "    \"\"\"Training and validation logging results to tensorboard and stdout\n",
        "\n",
        "    Params\n",
        "    ------\n",
        "    output_directory (string): directory to save checkpoints\n",
        "    log_directory (string) directory to save tensorboard logs\n",
        "    checkpoint_path(string): checkpoint path\n",
        "    n_gpus (int): number of gpus\n",
        "    rank (int): rank of current gpu\n",
        "    hparams (object): comma separated list of \"name=value\" pairs.\n",
        "    \"\"\"\n",
        "    if hparams.distributed_run:\n",
        "        init_distributed(hparams, n_gpus, rank, group_name)\n",
        "\n",
        "    torch.manual_seed(hparams.seed)\n",
        "    torch.cuda.manual_seed(hparams.seed)\n",
        "\n",
        "    model = load_model(hparams)\n",
        "    learning_rate = hparams.learning_rate\n",
        "    optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate,\n",
        "                                 weight_decay=hparams.weight_decay)\n",
        "\n",
        "    if hparams.fp16_run:\n",
        "        from apex import amp\n",
        "        model, optimizer = amp.initialize(\n",
        "            model, optimizer, opt_level='O2')\n",
        "\n",
        "    if hparams.distributed_run:\n",
        "        model = apply_gradient_allreduce(model)\n",
        "\n",
        "    criterion = Tacotron2Loss()\n",
        "\n",
        "    logger = prepare_directories_and_logger(\n",
        "        output_directory, log_directory, rank)\n",
        "\n",
        "    train_loader, valset, collate_fn = prepare_dataloaders(hparams)\n",
        "\n",
        "    # Load checkpoint if one exists\n",
        "    iteration = 0\n",
        "    epoch_offset = 0\n",
        "    if checkpoint_path is not None and os.path.isfile(checkpoint_path):\n",
        "        if warm_start:\n",
        "            model = warm_start_model(\n",
        "                checkpoint_path, model, hparams.ignore_layers)\n",
        "        else:\n",
        "            model, optimizer, _learning_rate, iteration = load_checkpoint(\n",
        "                checkpoint_path, model, optimizer)\n",
        "            if hparams.use_saved_learning_rate:\n",
        "                learning_rate = _learning_rate\n",
        "            iteration += 1  # next iteration is iteration + 1\n",
        "            epoch_offset = max(0, int(iteration / len(train_loader)))\n",
        "    else:\n",
        "      os.path.isfile(\"/content/tacotron2/pretrained_model\")\n",
        "      %cd /dev/null\n",
        "      !/content/tacotron2/megadown.sh https://mega.nz/#!WXY3RILA!KyoGHtfB_sdhmLFoykG2lKWhh0GFdwMkk7OwAjpQHRo --o pretrained_model\n",
        "      %cd /content/tacotron2\n",
        "      model = warm_start_model(\"/content/tacotron2/pretrained_model\", model, hparams.ignore_layers)\n",
        "      # download LJSpeech pretrained model if no checkpoint already exists\n",
        "    \n",
        "    start_eposh = time.perf_counter()\n",
        "    learning_rate = 0.0\n",
        "    model.train()\n",
        "    is_overflow = False\n",
        "    # ================ MAIN TRAINNIG LOOP! ===================\n",
        "    for epoch in tqdm(range(epoch_offset, hparams.epochs)):\n",
        "        print(\"\\nStarting Epoch: {} Iteration: {}\".format(epoch, iteration))\n",
        "        start_eposh = time.perf_counter() # eposh is russian, not a typo\n",
        "        for i, batch in tqdm(enumerate(train_loader), total=len(train_loader)):\n",
        "            start = time.perf_counter()\n",
        "            if iteration < hparams.decay_start: learning_rate = hparams.A_\n",
        "            else: iteration_adjusted = iteration - hparams.decay_start; learning_rate = (hparams.A_*(e**(-iteration_adjusted/hparams.B_))) + hparams.C_\n",
        "            learning_rate = max(hparams.min_learning_rate, learning_rate) # output the largest number\n",
        "            for param_group in optimizer.param_groups:\n",
        "                param_group['lr'] = learning_rate\n",
        "\n",
        "            model.zero_grad()\n",
        "            x, y = model.parse_batch(batch)\n",
        "            y_pred = model(x)\n",
        "\n",
        "            loss = criterion(y_pred, y)\n",
        "            if hparams.distributed_run:\n",
        "                reduced_loss = reduce_tensor(loss.data, n_gpus).item()\n",
        "            else:\n",
        "                reduced_loss = loss.item()\n",
        "            if hparams.fp16_run:\n",
        "                with amp.scale_loss(loss, optimizer) as scaled_loss:\n",
        "                    scaled_loss.backward()\n",
        "            else:\n",
        "                loss.backward()\n",
        "\n",
        "            if hparams.fp16_run:\n",
        "                grad_norm = torch.nn.utils.clip_grad_norm_(\n",
        "                    amp.master_params(optimizer), hparams.grad_clip_thresh)\n",
        "                is_overflow = math.isnan(grad_norm)\n",
        "            else:\n",
        "                grad_norm = torch.nn.utils.clip_grad_norm_(\n",
        "                    model.parameters(), hparams.grad_clip_thresh)\n",
        "\n",
        "            optimizer.step()\n",
        "\n",
        "            if not is_overflow and rank == 0:\n",
        "                duration = time.perf_counter() - start\n",
        "                logger.log_training(\n",
        "                    reduced_loss, grad_norm, learning_rate, duration, iteration)\n",
        "                #print(\"Batch {} loss {:.6f} Grad Norm {:.6f} Time {:.6f}\".format(iteration, reduced_loss, grad_norm, duration), end='\\r', flush=True)\n",
        "\n",
        "            iteration += 1\n",
        "        validate(model, criterion, valset, iteration,\n",
        "                 hparams.batch_size, n_gpus, collate_fn, logger,\n",
        "                 hparams.distributed_run, rank, epoch, start_eposh, learning_rate)\n",
        "        save_checkpoint(model, optimizer, learning_rate, iteration, checkpoint_path)\n",
        "        if log_directory2 != None:\n",
        "            copy_tree(log_directory, log_directory2)\n",
        "def check_dataset(hparams):\n",
        "    from utils import load_wav_to_torch, load_filepaths_and_text\n",
        "    import os\n",
        "    import numpy as np\n",
        "    def check_arr(filelist_arr):\n",
        "        for i, file in enumerate(filelist_arr):\n",
        "            if len(file) > 2:\n",
        "                print(\"|\".join(file), \"\\nhas multiple '|', this may not be an error.\")\n",
        "            if hparams.load_mel_from_disk and '.wav' in file[0]:\n",
        "                print(\"[WARNING]\", file[0], \" in filelist while expecting .npy .\")\n",
        "            else:\n",
        "                if not hparams.load_mel_from_disk and '.npy' in file[0]:\n",
        "                    print(\"[WARNING]\", file[0], \" in filelist while expecting .wav .\")\n",
        "            if (not os.path.exists(file[0])):\n",
        "                print(\"|\".join(file), \"\\n[WARNING] does not exist.\")\n",
        "            if len(file[1]) < 3:\n",
        "                print(\"|\".join(file), \"\\n[info] has no/very little text.\")\n",
        "            if not ((file[1].strip())[-1] in r\"!?,.;:\"):\n",
        "                print(\"|\".join(file), \"\\n[info] has no ending punctuation.\")\n",
        "            mel_length = 1\n",
        "            if hparams.load_mel_from_disk and '.npy' in file[0]:\n",
        "                melspec = torch.from_numpy(np.load(file[0], allow_pickle=True))\n",
        "                mel_length = melspec.shape[1]\n",
        "            if mel_length == 0:\n",
        "                print(\"|\".join(file), \"\\n[WARNING] has 0 duration.\")\n",
        "    print(\"Checking Training Files\")\n",
        "    audiopaths_and_text = load_filepaths_and_text(hparams.training_files) # get split lines from training_files text file.\n",
        "    check_arr(audiopaths_and_text)\n",
        "    print(\"Checking Validation Files\")\n",
        "    audiopaths_and_text = load_filepaths_and_text(hparams.validation_files) # get split lines from validation_files text file.\n",
        "    check_arr(audiopaths_and_text)\n",
        "    print(\"Finished Checking\")\n",
        "\n",
        "warm_start=False#sorry bout that\n",
        "n_gpus=1\n",
        "rank=0\n",
        "group_name=None\n",
        "\n",
        "# ---- DEFAULT PARAMETERS DEFINED HERE ----\n",
        "hparams = create_hparams()\n",
        "model_filename = 'current_model'\n",
        "hparams.training_files = \"filelists/clipper_train_filelist.txt\"\n",
        "hparams.validation_files = \"filelists/clipper_val_filelist.txt\"\n",
        "#hparams.use_mmi=True,          # not used in this notebook\n",
        "#hparams.use_gaf=True,          # not used in this notebook\n",
        "#hparams.max_gaf=0.5,           # not used in this notebook\n",
        "#hparams.drop_frame_rate = 0.2  # not used in this notebook\n",
        "hparams.p_attention_dropout=0.1\n",
        "hparams.p_decoder_dropout=0.1\n",
        "hparams.decay_start = 15000\n",
        "hparams.A_ = 5e-4\n",
        "hparams.B_ = 8000\n",
        "hparams.C_ = 0\n",
        "hparams.min_learning_rate = 1e-5\n",
        "generate_mels = True\n",
        "hparams.show_alignments = True\n",
        "alignment_graph_height = 600\n",
        "alignment_graph_width = 1000\n",
        "hparams.batch_size = 32\n",
        "hparams.load_mel_from_disk = True\n",
        "hparams.ignore_layers = []\n",
        "hparams.epochs = 10000\n",
        "torch.backends.cudnn.enabled = hparams.cudnn_enabled\n",
        "torch.backends.cudnn.benchmark = hparams.cudnn_benchmark\n",
        "output_directory = '/content/drive/My Drive/colab/outdir' # Location to save Checkpoints\n",
        "log_directory = '/content/tacotron2/logs' # Location to save Log files locally\n",
        "log_directory2 = '/content/drive/My Drive/colab/logs' # Location to copy log files (done at the end of each epoch to cut down on I/O)\n",
        "checkpoint_path = output_directory+(r'/')+model_filename\n",
        "\n",
        "# ---- Replace .wav with .npy in filelists ----\n",
        "!sed -i -- 's,.wav|,.npy|,g' filelists/*.txt\n",
        "!sed -i -- 's,.wav|,.npy|,g' {hparams.training_files}\n",
        "!sed -i -- 's,.wav|,.npy|,g' {hparams.validation_files}\n",
        "# ---- Replace .wav with .npy in filelists ----\n",
        "\n",
        "%cd /content/tacotron2\n",
        "\n",
        "data_path = 'wavs'\n",
        "!mkdir {data_path}"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "cellView": "form",
        "id": "kcfH75Dcw_uk"
      },
      "outputs": [],
      "source": [
        "#@markdown <b><font color=\"pink\"> Optional: \n",
        "\n",
        "#@markdown Revert all patches and reset \n",
        "#@markdown <b><font color=\"red\"> (DO NOT RUN)\n",
        "%cd /content/tacotron2/\n",
        "!git reset --hard HEAD"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2hQ3F8uZbTs8"
      },
      "source": [
        "### **Optional:** Run the cell below 👇 if you want to load another dataset and clear/remove the current one "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "cellView": "form",
        "id": "54KO-O3PbGGf"
      },
      "outputs": [],
      "source": [
        "#@markdown <b><font color=\"pink\" size=\"+2\"> Clean tacotron2 dataset\n",
        "!rm -rf /content/tacotron2/wavs/*\n",
        "!rm -rf /content/tacotron2/filelists/*"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OapsePIJBZq8"
      },
      "source": [
        "# **Training** "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "cellView": "form",
        "id": "UP1trdpN_jV6"
      },
      "outputs": [],
      "source": [
        "#@markdown ## <b><font color=\"pink\" size=\"+2\">  **Adjust model parameters**\n",
        "#@markdown ---\n",
        "\n",
        "#@markdown #### Your desired model name:\n",
        "\n",
        "model_filename = 'test' #@param {type: \"string\"}\n",
        "\n",
        "#@markdown #### Upload your transcription / text to tacotron2/filelists and right click -> copy path:\n",
        "Training_file = \"filelists/list.txt\" #@param {type: \"string\"}\n",
        "hparams.training_files = Training_file\n",
        "hparams.validation_files = Training_file\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "# hparams to Tune\n",
        "#hparams.use_mmi=True,          # not used in this notebook\n",
        "#hparams.use_gaf=True,          # not used in this notebook\n",
        "#hparams.max_gaf=0.5,           # not used in this notebook\n",
        "#hparams.drop_frame_rate = 0.2  # not used in this notebook\n",
        "hparams.p_attention_dropout=0.1\n",
        "hparams.p_decoder_dropout=0.1\n",
        "\n",
        "# Learning Rate             # https://www.desmos.com/calculator/ptgcz4vzsw / http://boards.4channel.org/mlp/thread/34778298#p34789030\n",
        "hparams.decay_start = 15000         # wait till decay_start to start decaying learning rate\n",
        "\n",
        "#@markdown #### Lower learning rates will take more time but will lead to more accurate results:\n",
        "# Start/Max Learning Rate  \n",
        "hparams.A_ = 5e-4 #@param [\"3e-6\", \"1e-5\", \"1e-4\", \"5e-4\", \"1e-3\"] {type:\"raw\", allow-input: true}              \n",
        "hparams.B_ = 8000                   # Decay Rate\n",
        "hparams.C_ = 0                      # Shift learning rate equation by this value\n",
        "hparams.min_learning_rate = 1e-5    # Min Learning Rate\n",
        "\n",
        "# Quality of Life\n",
        "generate_mels = True\n",
        "hparams.show_alignments = True\n",
        "alignment_graph_height = 600\n",
        "alignment_graph_width = 1000\n",
        "\n",
        "#@markdown #### Your batch size, lower if you don't have enough ram:\n",
        "\n",
        "hparams.batch_size =  14#@param {type: \"integer\"}\n",
        "hparams.load_mel_from_disk = True\n",
        "hparams.ignore_layers = [] # Layers to reset (None by default, other than foreign languages this param can be ignored)\n",
        "use_cmudict = True #@param {type:\"boolean\"}\n",
        "#@markdown #### Your total epochs to train to. Not recommended to change:\n",
        "\n",
        "##@markdown #### Amount of epochs before stopping, preferably a very high amount to not stop.\n",
        "hparams.epochs =  250#@param {type: \"integer\"}\n",
        "\n",
        "torch.backends.cudnn.enabled = hparams.cudnn_enabled\n",
        "torch.backends.cudnn.benchmark = hparams.cudnn_benchmark\n",
        "\n",
        "#@markdown #### Where to save your model when training:\n",
        "output_directory = '/content/drive/MyDrive/colab/outdir' #@param {type: \"string\"}\n",
        "log_directory = '/content/tacotron2/logs' # Location to save Log files locally\n",
        "log_directory2 = '/content/drive/My Drive/colab/logs' # Location to copy log files (done at the end of each epoch to cut down on I/O)\n",
        "checkpoint_path = output_directory+(r'/')+model_filename\n",
        "\n",
        "##@markdown #### Train the model from scratch? (If yes, then uncheck the box below):\n",
        "#warm_start=True #@param {type:\"boolean\"}\n",
        "\n",
        "\n",
        "#@markdown ---\n",
        "hparams.text_cleaners=[\"english_cleaners\"] + ([\"cmudict_cleaners\"] if use_cmudict is True else [])\n",
        "\n",
        "\n",
        "#@markdown Note:-\n",
        "\n",
        "#@markdown - The learning_rate value is ordered from smallest to largest, top to bottom.\n",
        "\n",
        "#@markdown - The smaller the \"learning rates\" value is, the longer it will take to train the model, but the more accurate the results will be.\n",
        "\n",
        "#@markdown ___\n",
        "\n",
        "#@markdown Todo:-\n",
        "#@markdown - Disable warm_start\n",
        "#@markdown - Add tensorboard training monitor\n",
        "\n",
        "#@markdown ___\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "cellView": "form",
        "id": "A2-xc6EcACWc"
      },
      "outputs": [],
      "source": [
        "#@markdown ## <b><font color=\"pink\" size=\"+2\">  **Convert WAV files into Mel-Spectrograms (Run only once)**\n",
        "\n",
        "if generate_mels:\n",
        "    create_mels()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "pDEiJHqJbLmY",
        "cellView": "form"
      },
      "outputs": [],
      "source": [
        "#@markdown ## <b><font color=\"pink\" size=\"+2\"> **Check the working cmudict patch**\n",
        "%cd /content/tacotron2/\n",
        "import text\n",
        "for i in range(10):\n",
        "  print(text.sequence_to_text(text.text_to_sequence(\"We must capture an Earth creature, K 9, and return it back with us to Mars.\", [\"cmudict_cleaners\", \"english_cleaners\"])))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "cellView": "form",
        "id": "j_ATUV9uWHwx"
      },
      "outputs": [],
      "source": [
        "#@markdown ## <b><font color=\"pink\" size=\"+2\"> **Check for missing files**\n",
        "\n",
        "# ---- Replace .wav with .npy in filelists ----\n",
        "!sed -i -- 's,.wav|,.npy|,g' {hparams.training_files}; sed -i -- 's,.wav|,.npy|,g' {hparams.validation_files}\n",
        "\n",
        "check_dataset(hparams)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "cellView": "form",
        "id": "iqIk1dIuCqhx"
      },
      "outputs": [],
      "source": [
        "#@markdown ## <b><font color=\"pink\" size=\"+2\">  **Begin training**\n",
        "\n",
        "print('FP16 Run:', hparams.fp16_run)\n",
        "print('Dynamic Loss Scaling:', hparams.dynamic_loss_scaling)\n",
        "print('Distributed Run:', hparams.distributed_run)\n",
        "print('cuDNN Enabled:', hparams.cudnn_enabled)\n",
        "print('cuDNN Benchmark:', hparams.cudnn_benchmark)\n",
        "train(output_directory, log_directory, checkpoint_path,\n",
        "      warm_start, n_gpus, rank, group_name, hparams, log_directory2)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "h4ql3OWWPKZv"
      },
      "source": [
        "JS to prevent idle timeout:\n",
        "\n",
        "Press F12 OR CTRL + SHIFT + I OR right click on this website -> inspect.\n",
        "Then click on the console tab and paste in the following code.\n",
        "\n",
        "```javascript\n",
        "function ClickConnect(){\n",
        "console.log(\"Working\");\n",
        "document.querySelector(\"colab-toolbar-button#connect\").click()\n",
        "}\n",
        "setInterval(ClickConnect,60000)\n",
        "```"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "collapsed_sections": [
        "iQBUZTuMogui",
        "2hQ3F8uZbTs8"
      ],
      "name": "bkp_FakeYou_Tacotron_2_(w_ARPAbet)_Training.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}