tutorial/1. Forced Aligner with Wav2vec2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "fff23186",
   "metadata": {},
   "source": [
    "# English(LJSpeech-1.1)\n",
    "\n",
    "- This is a tutorial to make `Forced Aligned Textfile` using [Wav2vec 2.0](https://arxiv.org/abs/2006.11477) released from [fairseq](https://github.com/pytorch/fairseq/blob/main/examples/wav2vec/README.md).\n",
    "- The original script to use wav2vec 2.0 as a forced aligner can be found in this [link](https://pytorch.org/tutorials/intermediate/forced_alignment_with_torchaudio_tutorial.html)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "6630c04f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from dataclasses import dataclass\n",
    "import torch\n",
    "import torchaudio\n",
    "import requests\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import IPython\n",
    "import pandas as pd\n",
    "import tgt\n",
    "from transformers import Wav2Vec2Processor, Wav2Vec2ForCTC\n",
    "from pathlib import Path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "9535b0e0",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Some weights of Wav2Vec2ForCTC were not initialized from the model checkpoint at facebook/wav2vec2-base-960h and are newly initialized: ['wav2vec2.masked_spec_embed']\n",
      "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Wav2Vec2ForCTC(\n",
       "  (wav2vec2): Wav2Vec2Model(\n",
       "    (feature_extractor): Wav2Vec2FeatureExtractor(\n",
       "      (conv_layers): ModuleList(\n",
       "        (0): Wav2Vec2GroupNormConvLayer(\n",
       "          (conv): Conv1d(1, 512, kernel_size=(10,), stride=(5,), bias=False)\n",
       "          (layer_norm): GroupNorm(512, 512, eps=1e-05, affine=True)\n",
       "        )\n",
       "        (1): Wav2Vec2NoLayerNormConvLayer(\n",
       "          (conv): Conv1d(512, 512, kernel_size=(3,), stride=(2,), bias=False)\n",
       "        )\n",
       "        (2): Wav2Vec2NoLayerNormConvLayer(\n",
       "          (conv): Conv1d(512, 512, kernel_size=(3,), stride=(2,), bias=False)\n",
       "        )\n",
       "        (3): Wav2Vec2NoLayerNormConvLayer(\n",
       "          (conv): Conv1d(512, 512, kernel_size=(3,), stride=(2,), bias=False)\n",
       "        )\n",
       "        (4): Wav2Vec2NoLayerNormConvLayer(\n",
       "          (conv): Conv1d(512, 512, kernel_size=(3,), stride=(2,), bias=False)\n",
       "        )\n",
       "        (5): Wav2Vec2NoLayerNormConvLayer(\n",
       "          (conv): Conv1d(512, 512, kernel_size=(2,), stride=(2,), bias=False)\n",
       "        )\n",
       "        (6): Wav2Vec2NoLayerNormConvLayer(\n",
       "          (conv): Conv1d(512, 512, kernel_size=(2,), stride=(2,), bias=False)\n",
       "        )\n",
       "      )\n",
       "    )\n",
       "    (feature_projection): Wav2Vec2FeatureProjection(\n",
       "      (layer_norm): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
       "      (projection): Linear(in_features=512, out_features=768, bias=True)\n",
       "      (dropout): Dropout(p=0.1, inplace=False)\n",
       "    )\n",
       "    (encoder): Wav2Vec2Encoder(\n",
       "      (pos_conv_embed): Wav2Vec2PositionalConvEmbedding(\n",
       "        (conv): Conv1d(768, 768, kernel_size=(128,), stride=(1,), padding=(64,), groups=16)\n",
       "        (padding): Wav2Vec2SamePadLayer()\n",
       "      )\n",
       "      (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "      (dropout): Dropout(p=0.1, inplace=False)\n",
       "      (layers): ModuleList(\n",
       "        (0): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (1): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (2): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (3): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (4): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (5): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (6): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (7): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (8): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (9): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (10): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (11): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "      )\n",
       "    )\n",
       "  )\n",
       "  (dropout): Dropout(p=0.1, inplace=False)\n",
       "  (lm_head): Linear(in_features=768, out_features=32, bias=True)\n",
       ")"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## set device\n",
    "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
    "\n",
    "## load model\n",
    "processor = Wav2Vec2Processor.from_pretrained(\"facebook/wav2vec2-base-960h\")\n",
    "model = Wav2Vec2ForCTC.from_pretrained(\"facebook/wav2vec2-base-960h\")\n",
    "model.to(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "1a5175b3",
   "metadata": {},
   "outputs": [],
   "source": [
    "## get dict\n",
    "dictionary = processor.tokenizer.get_vocab()\n",
    "labels = [item for item, idx in dictionary.items()]\n",
    "\n",
    "## get sampling_rate\n",
    "w2v_sample_rate = processor.feature_extractor.sampling_rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "b5c5c488",
   "metadata": {},
   "outputs": [],
   "source": [
    "## set meta and audio data path of LJSpeech-1.1\n",
    "data_path = '/code/gitRepo/data/LJSpeech-1.1/wavs'\n",
    "script_path = '/code/gitRepo/data/LJSpeech-1.1/metadata.csv'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "5747dd1b",
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(script_path, 'r') as f:\n",
    "    scripts = f.readlines()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "b2065960",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['LJ001-0002',\n",
       " 'in being comparatively modern.',\n",
       " 'in being comparatively modern.']"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "script_index = 1\n",
    "scripts[script_index].strip().split('|')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "f6b0bcf0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[audio_path] /code/gitRepo/data/LJSpeech-1.1/wavs/LJ001-0002.wav\n",
      "[transcript] in being comparatively modern.\n"
     ]
    }
   ],
   "source": [
    "item = scripts[script_index].strip().split('|')\n",
    "audio_path = os.path.join(data_path, item[0])\n",
    "if Path(audio_path).suffix == '':\n",
    "    audio_path = '{}.wav'.format(audio_path)\n",
    "transcript = item[2]\n",
    "print('[audio_path]', audio_path)\n",
    "print('[transcript]', transcript)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "50127415",
   "metadata": {},
   "outputs": [],
   "source": [
    "## get probs from wav2vec2\n",
    "with torch.inference_mode():\n",
    "    waveform, sample_rate = torchaudio.load(audio_path)\n",
    "\n",
    "    ## only use mono\n",
    "    if waveform.size(0) > 1:\n",
    "        waveform = waveform[0, :].view(1, -1)\n",
    "\n",
    "    if sample_rate != w2v_sample_rate:\n",
    "        waveform = torchaudio.transforms.Resample(sample_rate, w2v_sample_rate)(waveform)\n",
    "        sample_rate = w2v_sample_rate\n",
    "\n",
    "    emissions = model(waveform.to(device))\n",
    "    emissions = torch.log_softmax(emissions.logits, dim=-1)\n",
    "\n",
    "emission = emissions[0].cpu().detach()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "f942fdbc",
   "metadata": {},
   "outputs": [],
   "source": [
    "@dataclass\n",
    "class Segment:\n",
    "    label: str\n",
    "    start: int\n",
    "    end: int\n",
    "    score: float\n",
    "\n",
    "    def __repr__(self):\n",
    "        return f\"{self.label}\\t({self.score:4.2f}): [{self.start:5d}, {self.end:5d})\"\n",
    "\n",
    "    @property\n",
    "    def length(self):\n",
    "        return self.end - self.start\n",
    "    \n",
    "@dataclass\n",
    "class Point:\n",
    "    token_index: int\n",
    "    time_index: int\n",
    "    score: float\n",
    "\n",
    "\n",
    "def get_trellis(emission, tokens, blank_id=0):\n",
    "    num_frame = emission.size(0)\n",
    "    num_tokens = len(tokens)\n",
    "\n",
    "    # Trellis has extra diemsions for both time axis and tokens.\n",
    "    # The extra dim for tokens represents <SoS> (start-of-sentence)\n",
    "    # The extra dim for time axis is for simplification of the code.\n",
    "    trellis = torch.full((num_frame+1, num_tokens+1), -float('inf'))\n",
    "    trellis[:, 0] = 0\n",
    "    for t in range(num_frame):\n",
    "        trellis[t+1, 1:] = torch.maximum(\n",
    "            # Score for staying at the same token\n",
    "            trellis[t, 1:] + emission[t, blank_id],\n",
    "            # Score for changing to the next token\n",
    "            trellis[t, :-1] + emission[t, tokens],\n",
    "        )\n",
    "    return trellis\n",
    "\n",
    "def backtrack(trellis, emission, tokens, blank_id=0):\n",
    "    # Note:\n",
    "    # j and t are indices for trellis, which has extra dimensions\n",
    "    # for time and tokens at the beginning.\n",
    "    # When refering to time frame index `T` in trellis,\n",
    "    # the corresponding index in emission is `T-1`.\n",
    "    # Similarly, when refering to token index `J` in trellis,\n",
    "    # the corresponding index in transcript is `J-1`.\n",
    "    j = trellis.size(1) - 1\n",
    "    t_start = torch.argmax(trellis[:, j]).item()\n",
    "\n",
    "    path = []\n",
    "    for t in range(t_start, 0, -1):\n",
    "        # 1. Figure out if the current position was stay or change\n",
    "        # Note (again):\n",
    "        # `emission[J-1]` is the emission at time frame `J` of trellis dimension.\n",
    "        # Score for token staying the same from time frame J-1 to T.\n",
    "        stayed = trellis[t-1, j] + emission[t-1, blank_id]\n",
    "        # Score for token changing from C-1 at T-1 to J at T.\n",
    "        changed = trellis[t-1, j-1] + emission[t-1, tokens[j-1]]\n",
    "\n",
    "        # 2. Store the path with frame-wise probability.\n",
    "        prob = emission[t-1, tokens[j-1] if changed > stayed else 0].exp().item()\n",
    "        # Return token index and time index in non-trellis coordinate.\n",
    "        path.append(Point(j-1, t-1, prob))\n",
    "\n",
    "        # 3. Update the token\n",
    "        if changed > stayed:\n",
    "            j -= 1\n",
    "            if j == 0:\n",
    "                break\n",
    "    else:\n",
    "        raise ValueError('Failed to align')\n",
    "    return path[::-1]\n",
    "\n",
    "def merge_repeats(path, transcript):\n",
    "    \n",
    "    i1, i2 = 0, 0\n",
    "    segments = []\n",
    "    while i1 < len(path):\n",
    "        while i2 < len(path) and path[i1].token_index == path[i2].token_index:\n",
    "            i2 += 1\n",
    "        score = sum(path[k].score for k in range(i1, i2)) / (i2 - i1)\n",
    "        segments.append(Segment(transcript[path[i1].token_index], path[i1].time_index, path[i2-1].time_index + 1, score))\n",
    "        i1 = i2\n",
    "    return segments\n",
    "\n",
    "\n",
    "def plot_alignments(trellis, segments, word_segments, waveform):\n",
    "    trellis_with_path = trellis.clone()\n",
    "    for i, seg in enumerate(segments):\n",
    "        if seg.label != '|':\n",
    "            trellis_with_path[seg.start+1:seg.end+1, i+1] = float('nan')\n",
    "\n",
    "    fig, [ax1, ax2] = plt.subplots(2, 1, figsize=(16, 12.5))\n",
    "\n",
    "    ax1.imshow(trellis_with_path[1:, 1:].T, origin='lower')\n",
    "    ax1.set_xticks([])\n",
    "    ax1.set_yticks([])\n",
    "\n",
    "    for word in word_segments:\n",
    "        ax1.axvline(word.start - 0.5)\n",
    "        ax1.axvline(word.end - 0.5)\n",
    "\n",
    "    for i, seg in enumerate(segments):\n",
    "        if seg.label != '|':\n",
    "            ax1.annotate(seg.label, (seg.start, i + 0.3))\n",
    "            ax1.annotate(f'{seg.score:.2f}', (seg.start , i + 4), fontsize=8)\n",
    "\n",
    "  # The original waveform\n",
    "    ratio = waveform.size(0) / (trellis.size(0) - 1)\n",
    "    ax2.plot(waveform)\n",
    "    for word in word_segments:\n",
    "        x0 = ratio * word.start\n",
    "        x1 = ratio * word.end\n",
    "        ax2.axvspan(x0, x1, alpha=0.1, color='red')\n",
    "        ax2.annotate(f'{word.score:.2f}', (x0, 0.8))\n",
    "\n",
    "    for seg in segments:\n",
    "        if seg.label != '|':\n",
    "            ax2.annotate(seg.label, (seg.start * ratio, 0.9))\n",
    "    xticks = ax2.get_xticks()\n",
    "    plt.xticks(xticks, xticks / w2v_sample_rate)\n",
    "    ax2.set_xlabel('time [second]')\n",
    "    ax2.set_yticks([])\n",
    "    ax2.set_ylim(-1.0, 1.0)\n",
    "    ax2.set_xlim(0, waveform.size(-1))\n",
    "    plt.tight_layout()\n",
    "    plt.show()\n",
    "    \n",
    "# Merge words\n",
    "def merge_words(segments, separator='|'):\n",
    "    words = []\n",
    "    i1, i2 = 0, 0\n",
    "    while i1 < len(segments):\n",
    "        if i2 >= len(segments) or segments[i2].label == separator:\n",
    "            if i1 != i2:\n",
    "                segs = segments[i1:i2]\n",
    "                word = ''.join([seg.label for seg in segs])\n",
    "                score = sum(seg.score * seg.length for seg in segs) / sum(seg.length for seg in segs)\n",
    "                words.append(Segment(word, segments[i1].start, segments[i2-1].end, score))\n",
    "            i1 = i2 + 1\n",
    "            i2 = i1\n",
    "        else:\n",
    "            i2 += 1\n",
    "    return words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "c4497b4b",
   "metadata": {},
   "outputs": [],
   "source": [
    "## make tokens\n",
    "tokens = list()\n",
    "converted_items = list()\n",
    "new_transcript = '|' + transcript.replace(' ', '|').upper() + '|'\n",
    "#new_transcript = transcript.replace(' ', '|').upper()\n",
    "for c in new_transcript:\n",
    "    if c not in dictionary:\n",
    "        converted_items.append([len(tokens)-1, c])\n",
    "    else:\n",
    "        tokens.append(dictionary[c])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "58539ad0",
   "metadata": {},
   "outputs": [],
   "source": [
    "trellis = get_trellis(emission, tokens)\n",
    "back_path = backtrack(trellis, emission, tokens)\n",
    "segments = merge_repeats(back_path, [labels[token] for token in tokens])\n",
    "word_segments = merge_words(segments)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "dc05c9b0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x900 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_alignments(trellis, segments, word_segments, waveform[0],)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6cea879c",
   "metadata": {},
   "source": [
    "# Korean(KSS ver4)\n",
    "\n",
    "- This is a tutorial to make `Forced Aligned Textfile` using [Wav2vec 2.0](https://arxiv.org/abs/2006.11477) released from [fairseq](https://github.com/pytorch/fairseq/blob/main/examples/wav2vec/README.md).\n",
    "- The original script to use wav2vec 2.0 as a forced aligner can be found in this [link](https://pytorch.org/tutorials/intermediate/forced_alignment_with_torchaudio_tutorial.html)\n",
    "- To use Korean Wav2vec 2.0, you need to download trained model from this [link](https://drive.google.com/file/d/1BBYDTaBS0co7_VgaRowqk8NvFGA5bBgF/view?usp=sharing) which is trained by [JoungheeKim](https://github.com/JoungheeKim) using [Aihub Korean Dialog corpus](https://aihub.or.kr/aidata/105).\n",
    "- After downloading the file, you must unzip it to use the model. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "70d47bc4",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from dataclasses import dataclass\n",
    "import torch\n",
    "import torchaudio\n",
    "import requests\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import IPython\n",
    "import pandas as pd\n",
    "import tgt\n",
    "from transformers import Wav2Vec2Processor, Wav2Vec2ForCTC\n",
    "from pathlib import Path\n",
    "import unicodedata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "24363498",
   "metadata": {},
   "outputs": [],
   "source": [
    "## make sure you have korean font to visualize the output\n",
    "import matplotlib.font_manager as fonm\n",
    "font_list = [font.name for font in fonm.fontManager.ttflist]\n",
    "plt.rcParams[\"font.family\"] = 'NanumGothic'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "6e717c65",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Wav2Vec2ForCTC(\n",
       "  (wav2vec2): Wav2Vec2Model(\n",
       "    (feature_extractor): Wav2Vec2FeatureExtractor(\n",
       "      (conv_layers): ModuleList(\n",
       "        (0): Wav2Vec2GroupNormConvLayer(\n",
       "          (conv): Conv1d(1, 512, kernel_size=(10,), stride=(5,), bias=False)\n",
       "          (layer_norm): GroupNorm(512, 512, eps=1e-05, affine=True)\n",
       "        )\n",
       "        (1): Wav2Vec2NoLayerNormConvLayer(\n",
       "          (conv): Conv1d(512, 512, kernel_size=(3,), stride=(2,), bias=False)\n",
       "        )\n",
       "        (2): Wav2Vec2NoLayerNormConvLayer(\n",
       "          (conv): Conv1d(512, 512, kernel_size=(3,), stride=(2,), bias=False)\n",
       "        )\n",
       "        (3): Wav2Vec2NoLayerNormConvLayer(\n",
       "          (conv): Conv1d(512, 512, kernel_size=(3,), stride=(2,), bias=False)\n",
       "        )\n",
       "        (4): Wav2Vec2NoLayerNormConvLayer(\n",
       "          (conv): Conv1d(512, 512, kernel_size=(3,), stride=(2,), bias=False)\n",
       "        )\n",
       "        (5): Wav2Vec2NoLayerNormConvLayer(\n",
       "          (conv): Conv1d(512, 512, kernel_size=(2,), stride=(2,), bias=False)\n",
       "        )\n",
       "        (6): Wav2Vec2NoLayerNormConvLayer(\n",
       "          (conv): Conv1d(512, 512, kernel_size=(2,), stride=(2,), bias=False)\n",
       "        )\n",
       "      )\n",
       "    )\n",
       "    (feature_projection): Wav2Vec2FeatureProjection(\n",
       "      (layer_norm): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
       "      (projection): Linear(in_features=512, out_features=768, bias=True)\n",
       "      (dropout): Dropout(p=0.0, inplace=False)\n",
       "    )\n",
       "    (encoder): Wav2Vec2Encoder(\n",
       "      (pos_conv_embed): Wav2Vec2PositionalConvEmbedding(\n",
       "        (conv): Conv1d(768, 768, kernel_size=(128,), stride=(1,), padding=(64,), groups=16)\n",
       "        (padding): Wav2Vec2SamePadLayer()\n",
       "      )\n",
       "      (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "      (dropout): Dropout(p=0.1, inplace=False)\n",
       "      (layers): ModuleList(\n",
       "        (0): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (1): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (2): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (3): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (4): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (5): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (6): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (7): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (8): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (9): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (10): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "        (11): Wav2Vec2EncoderLayer(\n",
       "          (attention): Wav2Vec2Attention(\n",
       "            (k_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (v_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (q_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "            (out_proj): Linear(in_features=768, out_features=768, bias=True)\n",
       "          )\n",
       "          (dropout): Dropout(p=0.1, inplace=False)\n",
       "          (layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "          (feed_forward): Wav2Vec2FeedForward(\n",
       "            (intermediate_dropout): Dropout(p=0.1, inplace=False)\n",
       "            (intermediate_dense): Linear(in_features=768, out_features=3072, bias=True)\n",
       "            (output_dense): Linear(in_features=3072, out_features=768, bias=True)\n",
       "            (output_dropout): Dropout(p=0.1, inplace=False)\n",
       "          )\n",
       "          (final_layer_norm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
       "        )\n",
       "      )\n",
       "    )\n",
       "  )\n",
       "  (dropout): Dropout(p=0.1, inplace=False)\n",
       "  (lm_head): Linear(in_features=768, out_features=108, bias=True)\n",
       ")"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## set device\n",
    "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
    "\n",
    "## load model\n",
    "processor = Wav2Vec2Processor.from_pretrained(\"korean_wav2vec2\")\n",
    "model = Wav2Vec2ForCTC.from_pretrained(\"korean_wav2vec2\")\n",
    "model.to(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "023b7986",
   "metadata": {},
   "outputs": [],
   "source": [
    "## get dict\n",
    "dictionary = processor.tokenizer.get_vocab()\n",
    "labels = [item for item, idx in dictionary.items()]\n",
    "\n",
    "## get sampling_rate\n",
    "w2v_sample_rate = processor.feature_extractor.sampling_rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "d1419ca2",
   "metadata": {},
   "outputs": [],
   "source": [
    "## set meta and audio data path of KSS dataset\n",
    "data_path = '/code/gitRepo/data/kss'\n",
    "script_path = '/code/gitRepo/data/kss/transcript.v.1.4.txt'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "117d269c",
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(script_path, 'r') as f:\n",
    "    scripts = f.readlines()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "b4d4bb21",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['1/1_0001.wav',\n",
       " '그녀의 사랑을 얻기 위해 애썼지만 헛수고였다.',\n",
       " '그녀의 사랑을 얻기 위해 애썼지만 헛수고였다.',\n",
       " '그녀의 사랑을 얻기 위해 애썼지만 헛수고였다.',\n",
       " '4.0',\n",
       " 'I tried in vain to win her love.']"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "script_index = 1\n",
    "scripts[script_index].strip().split('|')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "deb10d1a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[audio_path] /code/gitRepo/data/kss/1/1_0001.wav\n",
      "[transcript] 그녀의 사랑을 얻기 위해 애썼지만 헛수고였다.\n"
     ]
    }
   ],
   "source": [
    "item = scripts[script_index].strip().split('|')\n",
    "audio_path = os.path.join(data_path, item[0])\n",
    "if Path(audio_path).suffix == '':\n",
    "    audio_path = '{}.wav'.format(audio_path)\n",
    "transcript = item[2]\n",
    "\n",
    "## normalize text for korean jamo\n",
    "transcript = unicodedata.normalize('NFKD', transcript.lower())\n",
    "print('[audio_path]', audio_path)\n",
    "print('[transcript]', transcript)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "bf73788e",
   "metadata": {},
   "outputs": [],
   "source": [
    "## get probs from wav2vec2\n",
    "with torch.inference_mode():\n",
    "    waveform, sample_rate = torchaudio.load(audio_path)\n",
    "\n",
    "    ## only use mono\n",
    "    if waveform.size(0) > 1:\n",
    "        waveform = waveform[0, :].view(1, -1)\n",
    "\n",
    "    if sample_rate != w2v_sample_rate:\n",
    "        waveform = torchaudio.transforms.Resample(sample_rate, w2v_sample_rate)(waveform)\n",
    "        sample_rate = w2v_sample_rate\n",
    "\n",
    "    emissions = model(waveform.to(device))\n",
    "    emissions = torch.log_softmax(emissions.logits, dim=-1)\n",
    "\n",
    "emission = emissions[0].cpu().detach()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "3b4f0901",
   "metadata": {},
   "outputs": [],
   "source": [
    "@dataclass\n",
    "class Segment:\n",
    "    label: str\n",
    "    start: int\n",
    "    end: int\n",
    "    score: float\n",
    "\n",
    "    def __repr__(self):\n",
    "        return f\"{self.label}\\t({self.score:4.2f}): [{self.start:5d}, {self.end:5d})\"\n",
    "\n",
    "    @property\n",
    "    def length(self):\n",
    "        return self.end - self.start\n",
    "    \n",
    "@dataclass\n",
    "class Point:\n",
    "    token_index: int\n",
    "    time_index: int\n",
    "    score: float\n",
    "\n",
    "\n",
    "def get_trellis(emission, tokens, blank_id=0):\n",
    "    num_frame = emission.size(0)\n",
    "    num_tokens = len(tokens)\n",
    "\n",
    "    # Trellis has extra diemsions for both time axis and tokens.\n",
    "    # The extra dim for tokens represents <SoS> (start-of-sentence)\n",
    "    # The extra dim for time axis is for simplification of the code.\n",
    "    trellis = torch.full((num_frame+1, num_tokens+1), -float('inf'))\n",
    "    trellis[:, 0] = 0\n",
    "    for t in range(num_frame):\n",
    "        trellis[t+1, 1:] = torch.maximum(\n",
    "            # Score for staying at the same token\n",
    "            trellis[t, 1:] + emission[t, blank_id],\n",
    "            # Score for changing to the next token\n",
    "            trellis[t, :-1] + emission[t, tokens],\n",
    "        )\n",
    "    return trellis\n",
    "\n",
    "def backtrack(trellis, emission, tokens, blank_id=0):\n",
    "    # Note:\n",
    "    # j and t are indices for trellis, which has extra dimensions\n",
    "    # for time and tokens at the beginning.\n",
    "    # When refering to time frame index `T` in trellis,\n",
    "    # the corresponding index in emission is `T-1`.\n",
    "    # Similarly, when refering to token index `J` in trellis,\n",
    "    # the corresponding index in transcript is `J-1`.\n",
    "    j = trellis.size(1) - 1\n",
    "    t_start = torch.argmax(trellis[:, j]).item()\n",
    "\n",
    "    path = []\n",
    "    for t in range(t_start, 0, -1):\n",
    "        # 1. Figure out if the current position was stay or change\n",
    "        # Note (again):\n",
    "        # `emission[J-1]` is the emission at time frame `J` of trellis dimension.\n",
    "        # Score for token staying the same from time frame J-1 to T.\n",
    "        stayed = trellis[t-1, j] + emission[t-1, blank_id]\n",
    "        # Score for token changing from C-1 at T-1 to J at T.\n",
    "        changed = trellis[t-1, j-1] + emission[t-1, tokens[j-1]]\n",
    "\n",
    "        # 2. Store the path with frame-wise probability.\n",
    "        prob = emission[t-1, tokens[j-1] if changed > stayed else 0].exp().item()\n",
    "        # Return token index and time index in non-trellis coordinate.\n",
    "        path.append(Point(j-1, t-1, prob))\n",
    "\n",
    "        # 3. Update the token\n",
    "        if changed > stayed:\n",
    "            j -= 1\n",
    "            if j == 0:\n",
    "                break\n",
    "    else:\n",
    "        raise ValueError('Failed to align')\n",
    "    return path[::-1]\n",
    "\n",
    "def merge_repeats(path, transcript):\n",
    "    \n",
    "    i1, i2 = 0, 0\n",
    "    segments = []\n",
    "    while i1 < len(path):\n",
    "        while i2 < len(path) and path[i1].token_index == path[i2].token_index:\n",
    "            i2 += 1\n",
    "        score = sum(path[k].score for k in range(i1, i2)) / (i2 - i1)\n",
    "        segments.append(Segment(transcript[path[i1].token_index], path[i1].time_index, path[i2-1].time_index + 1, score))\n",
    "        i1 = i2\n",
    "    return segments\n",
    "\n",
    "\n",
    "def plot_alignments(trellis, segments, word_segments, waveform):\n",
    "    trellis_with_path = trellis.clone()\n",
    "    for i, seg in enumerate(segments):\n",
    "        if seg.label != '|':\n",
    "            trellis_with_path[seg.start+1:seg.end+1, i+1] = float('nan')\n",
    "\n",
    "    fig, [ax1, ax2] = plt.subplots(2, 1, figsize=(16, 12.5))\n",
    "\n",
    "    ax1.imshow(trellis_with_path[1:, 1:].T, origin='lower')\n",
    "    ax1.set_xticks([])\n",
    "    ax1.set_yticks([])\n",
    "\n",
    "    for word in word_segments:\n",
    "        ax1.axvline(word.start - 0.5)\n",
    "        ax1.axvline(word.end - 0.5)\n",
    "\n",
    "    for i, seg in enumerate(segments):\n",
    "        if seg.label != '|':\n",
    "            ax1.annotate(seg.label, (seg.start, i + 0.3))\n",
    "            ax1.annotate(f'{seg.score:.2f}', (seg.start , i + 4), fontsize=8)\n",
    "\n",
    "  # The original waveform\n",
    "    ratio = waveform.size(0) / (trellis.size(0) - 1)\n",
    "    ax2.plot(waveform)\n",
    "    for word in word_segments:\n",
    "        x0 = ratio * word.start\n",
    "        x1 = ratio * word.end\n",
    "        ax2.axvspan(x0, x1, alpha=0.1, color='red')\n",
    "        ax2.annotate(f'{word.score:.2f}', (x0, 0.8))\n",
    "\n",
    "    for seg in segments:\n",
    "        if seg.label != '|':\n",
    "            ax2.annotate(seg.label, (seg.start * ratio, 0.9))\n",
    "    xticks = ax2.get_xticks()\n",
    "    plt.xticks(xticks, xticks / w2v_sample_rate)\n",
    "    ax2.set_xlabel('time [second]')\n",
    "    ax2.set_yticks([])\n",
    "    ax2.set_ylim(-1.0, 1.0)\n",
    "    ax2.set_xlim(0, waveform.size(-1))\n",
    "    plt.tight_layout()\n",
    "    plt.show()\n",
    "    \n",
    "# Merge words\n",
    "def merge_words(segments, separator='|'):\n",
    "    words = []\n",
    "    i1, i2 = 0, 0\n",
    "    while i1 < len(segments):\n",
    "        if i2 >= len(segments) or segments[i2].label == separator:\n",
    "            if i1 != i2:\n",
    "                segs = segments[i1:i2]\n",
    "                word = ''.join([seg.label for seg in segs])\n",
    "                score = sum(seg.score * seg.length for seg in segs) / sum(seg.length for seg in segs)\n",
    "                words.append(Segment(word, segments[i1].start, segments[i2-1].end, score))\n",
    "            i1 = i2 + 1\n",
    "            i2 = i1\n",
    "        else:\n",
    "            i2 += 1\n",
    "    return words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "be1f9384",
   "metadata": {},
   "outputs": [],
   "source": [
    "## make tokens\n",
    "tokens = list()\n",
    "converted_items = list()\n",
    "new_transcript = '|' + transcript.replace(' ', '|').upper() + '|'\n",
    "#new_transcript = transcript.replace(' ', '|').upper()\n",
    "for c in new_transcript:\n",
    "    if c not in dictionary:\n",
    "        converted_items.append([len(tokens)-1, c])\n",
    "    else:\n",
    "        tokens.append(dictionary[c])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "9c6a8ead",
   "metadata": {},
   "outputs": [],
   "source": [
    "trellis = get_trellis(emission, tokens)\n",
    "back_path = backtrack(trellis, emission, tokens)\n",
    "segments = merge_repeats(back_path, [labels[token] for token in tokens])\n",
    "segments = segments\n",
    "word_segments = merge_words(segments)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "802239b3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x900 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_alignments(trellis, segments, word_segments, waveform[0],)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}