Audio/golos_asr_400.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "oGrdjDX-QCP5",
   "metadata": {
    "id": "oGrdjDX-QCP5",
    "tags": []
   },
   "outputs": [],
   "source": [
    "# %%time\n",
    "# !tar -xf golos_100hours.tar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "1198e47d-4ec1-4574-8bca-412046481449",
   "metadata": {
    "id": "1198e47d-4ec1-4574-8bca-412046481449",
    "tags": []
   },
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torchaudio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "e859e6a0-e3c5-4e05-a57e-29016ce2feab",
   "metadata": {
    "id": "e859e6a0-e3c5-4e05-a57e-29016ce2feab",
    "tags": []
   },
   "outputs": [],
   "source": [
    "import torch.nn.functional as F\n",
    "from torch.nn.utils.rnn import pad_sequence\n",
    "from torch.utils.data import DataLoader, Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "692883cd-a62c-4cd7-9510-5c77222e1d10",
   "metadata": {
    "id": "692883cd-a62c-4cd7-9510-5c77222e1d10",
    "tags": []
   },
   "outputs": [],
   "source": [
    "import logging\n",
    "import re\n",
    "import os\n",
    "import sys\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from IPython.display import clear_output\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "e1f467fa-4415-4b7b-b051-18e7f60cb7c7",
   "metadata": {
    "id": "e1f467fa-4415-4b7b-b051-18e7f60cb7c7",
    "tags": []
   },
   "outputs": [],
   "source": [
    "logging.basicConfig(\n",
    "    level=logging.INFO,\n",
    "    format=\"%(asctime)s - %(name)s - %(levelname)s - %(message)s\",\n",
    "    handlers=[logging.StreamHandler(sys.stdout)],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "7754b3f7-9c41-433b-aef0-ff0f08fd3997",
   "metadata": {
    "id": "7754b3f7-9c41-433b-aef0-ff0f08fd3997",
    "tags": []
   },
   "outputs": [],
   "source": [
    "class QuartzNet(nn.Module):\n",
    "    def __init__(self, R_repeat, in_channels, out_channels):\n",
    "        super(QuartzNet, self).__init__()\n",
    "\n",
    "        block_channels = [256, 256, 256, 512, 512, 512]\n",
    "        block_k = [33, 39, 51, 63, 75]\n",
    "\n",
    "        self.C1 = nn.Sequential(\n",
    "            nn.Conv1d(\n",
    "                in_channels,\n",
    "                in_channels,\n",
    "                kernel_size=33,\n",
    "                padding=16,\n",
    "                stride=2,\n",
    "                groups=in_channels,\n",
    "                bias=False,\n",
    "            ),\n",
    "            nn.Conv1d(in_channels, 256, kernel_size=1, stride=1, bias=False),\n",
    "            nn.BatchNorm1d(256, eps=0.001, momentum=0.1, affine=True),\n",
    "            nn.ReLU(),\n",
    "            nn.Dropout(p=0.2),\n",
    "        )\n",
    "\n",
    "        self.B = nn.ModuleList([])\n",
    "\n",
    "        for i in range(5):\n",
    "            num_in = block_channels[i]\n",
    "            num_out = block_channels[i + 1]\n",
    "            pad = block_k[i] // 2\n",
    "            k = block_k[i]\n",
    "\n",
    "            print(i, num_in, num_out, pad, k)\n",
    "\n",
    "            self.B.append(JasperBlock(num_in, num_out, k, pad))\n",
    "\n",
    "            for rep in range(R_repeat):\n",
    "                self.B.append(JasperBlock(num_out, num_out, k, pad))\n",
    "\n",
    "        self.C2 = nn.Sequential(\n",
    "            nn.Conv1d(\n",
    "                512, 512, kernel_size=87, padding=86, dilation=2, groups=512, bias=False\n",
    "            ),\n",
    "            nn.Conv1d(512, 512, kernel_size=1, stride=1, bias=False),\n",
    "            nn.BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True),\n",
    "            nn.ReLU(),\n",
    "            nn.Dropout(p=0.2),\n",
    "        )\n",
    "\n",
    "        self.C3 = nn.Sequential(\n",
    "            nn.Conv1d(512, 1024, kernel_size=1, stride=1, bias=False),\n",
    "            nn.BatchNorm1d(1024, eps=0.001, momentum=0.1, affine=True),\n",
    "            nn.ReLU(),\n",
    "            nn.Dropout(p=0.2),\n",
    "        )\n",
    "\n",
    "        self.C4 = nn.Conv1d(1024, out_channels, kernel_size=1, stride=1)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = self.C1(x)\n",
    "\n",
    "        for block in self.B:\n",
    "            x = block(x)\n",
    "\n",
    "        x = self.C2(x)\n",
    "        x = self.C3(x)\n",
    "        x = self.C4(x)\n",
    "\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "355fe27c-d4cb-4eb8-b9d3-638860fd2ebe",
   "metadata": {
    "id": "355fe27c-d4cb-4eb8-b9d3-638860fd2ebe",
    "tags": []
   },
   "outputs": [],
   "source": [
    "class JasperBlock(nn.Module):\n",
    "    def __init__(self, in_channels, out_channels, k, padding):\n",
    "        super(JasperBlock, self).__init__()\n",
    "\n",
    "        self.blocks = nn.Sequential(\n",
    "            ConvBlock(in_channels, out_channels, k, padding),\n",
    "            ConvBlock(out_channels, out_channels, k, padding),\n",
    "            ConvBlock(out_channels, out_channels, k, padding),\n",
    "            ConvBlock(out_channels, out_channels, k, padding),\n",
    "            ConvBlock(out_channels, out_channels, k, padding),\n",
    "        )\n",
    "\n",
    "        self.residual = nn.Sequential(\n",
    "            nn.Conv1d(\n",
    "                in_channels, out_channels, kernel_size=(1,), stride=[1], bias=False\n",
    "            ),\n",
    "            nn.BatchNorm1d(out_channels, eps=0.001, momentum=0.1, affine=True),\n",
    "        )\n",
    "\n",
    "        self.relu = nn.ReLU()\n",
    "        self.dropout = nn.Dropout(p=0.2)\n",
    "\n",
    "    def forward(self, x):\n",
    "        return self.dropout(self.relu(self.residual(x) + self.blocks(x)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "c77a14f0-b7a5-469b-880c-bb1826ad6c0d",
   "metadata": {
    "id": "c77a14f0-b7a5-469b-880c-bb1826ad6c0d",
    "tags": []
   },
   "outputs": [],
   "source": [
    "class ConvBlock(nn.Module):\n",
    "    def __init__(self, in_channels, out_channels, k, padding):\n",
    "        super(ConvBlock, self).__init__()\n",
    "\n",
    "        self.layers = nn.Sequential(\n",
    "            nn.Conv1d(\n",
    "                in_channels,\n",
    "                in_channels,\n",
    "                kernel_size=k,\n",
    "                stride=[1],\n",
    "                padding=(padding,),\n",
    "                dilation=[1],\n",
    "                groups=in_channels,\n",
    "                bias=False,\n",
    "            ),\n",
    "            nn.Conv1d(\n",
    "                in_channels, out_channels, kernel_size=(1,), stride=(1,), bias=False\n",
    "            ),\n",
    "            nn.BatchNorm1d(out_channels, eps=0.001, momentum=0.1, affine=True),\n",
    "            nn.ReLU(),\n",
    "            nn.Dropout(p=0.2),\n",
    "        )\n",
    "\n",
    "    def forward(self, x):\n",
    "        return self.layers(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "3155db4a-26b6-46d8-96ae-7e3bf26e276c",
   "metadata": {
    "id": "3155db4a-26b6-46d8-96ae-7e3bf26e276c",
    "tags": []
   },
   "outputs": [],
   "source": [
    "# files = (\n",
    "#     pd.read_json(path_or_buf=\"100hours.jsonl\", lines=True)\n",
    "#     .iloc[:75_000]\n",
    "#     .reset_index(drop=True)\n",
    "# )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "04a3a5d3-664b-41b1-b7bc-2c136b651e11",
   "metadata": {
    "id": "NR8cSgjpQSyc",
    "tags": []
   },
   "outputs": [],
   "source": [
    "# test_files = (\n",
    "#     pd.read_json(path_or_buf=\"100hours.jsonl\", lines=True)\n",
    "#     .iloc[75_000:]\n",
    "#     .reset_index(drop=True)\n",
    "# )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "254c453b-4bb4-46bb-a282-592ac4106b04",
   "metadata": {
    "id": "3155db4a-26b6-46d8-96ae-7e3bf26e276c",
    "tags": []
   },
   "outputs": [],
   "source": [
    "files = (\n",
    "    pd.read_json(path_or_buf=\"train_opus/manifest.jsonl\", lines=True)\n",
    ")\n",
    "files = files[files['duration'] > 1].sample(400_000).reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "89a8c9fb-1ee9-451f-baf7-abf91a1f9e33",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "444.93871824652786"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "files.duration.sum() / 3600"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "ca79577b-84aa-46fd-a55f-9c050d8e95d8",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_crowd = pd.read_json(path_or_buf=\"test_opus/crowd/manifest.jsonl\", lines=True)\n",
    "test_crowd['audio_filepath'] = 'crowd/' + test_crowd['audio_filepath']\n",
    "\n",
    "test_farfield = pd.read_json(path_or_buf=\"test_opus/farfield/manifest.jsonl\", lines=True)\n",
    "test_farfield['audio_filepath'] = 'farfield/' + test_farfield['audio_filepath']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "NR8cSgjpQSyc",
   "metadata": {
    "id": "NR8cSgjpQSyc",
    "tags": []
   },
   "outputs": [],
   "source": [
    "test_files = (\n",
    "    pd.concat(\n",
    "        [\n",
    "            test_crowd,\n",
    "            test_farfield,\n",
    "        ]\n",
    "    )\n",
    "    .reset_index(drop=True)\n",
    ")\n",
    "test_files = test_files[test_files['duration'] > 1].reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "0d5c9e1d-ddb6-4f86-94bc-ccaa5fa34440",
   "metadata": {
    "id": "0d5c9e1d-ddb6-4f86-94bc-ccaa5fa34440",
    "tags": []
   },
   "outputs": [],
   "source": [
    "chars = [\n",
    "    \" \",\n",
    "    \"а\",\n",
    "    \"б\",\n",
    "    \"в\",\n",
    "    \"г\",\n",
    "    \"д\",\n",
    "    \"е\",\n",
    "    \"ж\",\n",
    "    \"з\",\n",
    "    \"и\",\n",
    "    \"й\",\n",
    "    \"к\",\n",
    "    \"л\",\n",
    "    \"м\",\n",
    "    \"н\",\n",
    "    \"о\",\n",
    "    \"п\",\n",
    "    \"р\",\n",
    "    \"с\",\n",
    "    \"т\",\n",
    "    \"у\",\n",
    "    \"ф\",\n",
    "    \"х\",\n",
    "    \"ц\",\n",
    "    \"ч\",\n",
    "    \"ш\",\n",
    "    \"щ\",\n",
    "    \"ъ\",\n",
    "    \"ы\",\n",
    "    \"ь\",\n",
    "    \"э\",\n",
    "    \"ю\",\n",
    "    \"я\",\n",
    "]\n",
    "int_to_char = dict(enumerate(chars))\n",
    "char_to_int = {v: k for k, v in int_to_char.items()}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "7e6a800c-9bc5-485d-83b6-ffe6cfeed231",
   "metadata": {
    "id": "7e6a800c-9bc5-485d-83b6-ffe6cfeed231",
    "tags": []
   },
   "outputs": [],
   "source": [
    "class GolosDataset(Dataset):\n",
    "    def __init__(self, mode: str, files, spec_aug=None):\n",
    "        super(GolosDataset, self).__init__()\n",
    "\n",
    "        self.mode = mode\n",
    "        self.collection = files\n",
    "        self.sr = 16_000\n",
    "        self.spec_aug = spec_aug\n",
    "\n",
    "    def __len__(self):\n",
    "        return len(self.collection)\n",
    "\n",
    "    def __getitem__(self, item):\n",
    "        audio_path = self.collection[\"audio_filepath\"][item]\n",
    "        transcript = self.collection[\"text\"][item]\n",
    "\n",
    "        transcript_tensor = torch.tensor(\n",
    "            [char_to_int[i] for i in transcript], dtype=torch.long\n",
    "        )\n",
    "\n",
    "        if self.mode == \"train\":\n",
    "            waveform, sample_rate = torchaudio.load(\n",
    "                \"train_opus/\" + audio_path, normalize=True\n",
    "            )\n",
    "        else:\n",
    "            waveform, sample_rate = torchaudio.load(\n",
    "                \"test_opus/\" + audio_path, normalize=True\n",
    "            )\n",
    "        waveform = torchaudio.transforms.Resample(sample_rate, self.sr)(waveform)\n",
    "        \n",
    "        if self.spec_aug is None:\n",
    "            mel_spectrogram = torchaudio.transforms.MelSpectrogram(\n",
    "                n_mels=64,\n",
    "                sample_rate=self.sr,\n",
    "                n_fft=400,\n",
    "                power=2,\n",
    "                f_min=0.0,\n",
    "            )(waveform)\n",
    "        else:\n",
    "            # Convert to power spectrogram\n",
    "            spectrogram = torchaudio.transforms.Spectrogram(\n",
    "                n_fft=400,\n",
    "                power=2,\n",
    "            )(waveform)\n",
    "            # Apply SpecAugment\n",
    "            aug_spectrogram = self.spec_aug(spectrogram)\n",
    "            # Convert to mel-scale\n",
    "            mel_spectrogram = torchaudio.transforms.MelScale(\n",
    "                n_mels=64,\n",
    "                sample_rate=self.sr,\n",
    "                f_min=0.0,\n",
    "            )(aug_spectrogram)\n",
    "        \n",
    "        log_mel_spec = torch.log(torch.clamp(mel_spectrogram, min=1e-10))\n",
    "\n",
    "        input_length = log_mel_spec.shape[-1]\n",
    "        label_length = len(transcript_tensor)\n",
    "\n",
    "        return {\n",
    "            \"melspec\": log_mel_spec.squeeze(0),\n",
    "            \"text\": transcript_tensor,\n",
    "            \"input_length\": input_length,\n",
    "            \"label_length\": label_length,\n",
    "        }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "9edef4e7-8b59-4e99-a928-84ac0843efbc",
   "metadata": {},
   "outputs": [],
   "source": [
    "spec_aug = torch.nn.Sequential(\n",
    "    # torchaudio.transforms.TimeStretch(0.9, fixed_rate=True),\n",
    "    torchaudio.transforms.FrequencyMasking(freq_mask_param=20),\n",
    "    torchaudio.transforms.TimeMasking(time_mask_param=5),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "03108d4c-2487-4d51-9bad-09c6e7b99840",
   "metadata": {
    "id": "03108d4c-2487-4d51-9bad-09c6e7b99840",
    "tags": []
   },
   "outputs": [],
   "source": [
    "def count_parameters(model: nn.Module):\n",
    "    \"\"\"Количество обучаемых параметров модели\"\"\"\n",
    "    return sum(p.numel() for p in model.parameters() if p.requires_grad)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "aeb3d3ef-93af-4566-bd2c-183ac79c2f70",
   "metadata": {
    "id": "aeb3d3ef-93af-4566-bd2c-183ac79c2f70",
    "tags": []
   },
   "outputs": [],
   "source": [
    "train_dataset = GolosDataset(\"train\", files, spec_aug)\n",
    "test_dataset = GolosDataset(\"test\", test_files)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "47458051-07b0-4524-888d-e92917d4893c",
   "metadata": {
    "id": "47458051-07b0-4524-888d-e92917d4893c",
    "tags": []
   },
   "outputs": [],
   "source": [
    "data = train_dataset[1]\n",
    "log_mel_spec = data[\"melspec\"]\n",
    "transcript = data[\"text\"]\n",
    "input_length = data[\"input_length\"]\n",
    "label_length = data[\"label_length\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "d1ca50ff-888a-467c-93c8-fd17607274b7",
   "metadata": {
    "id": "47458051-07b0-4524-888d-e92917d4893c",
    "tags": []
   },
   "outputs": [],
   "source": [
    "data = test_dataset[1]\n",
    "log_mel_spec = data[\"melspec\"]\n",
    "transcript = data[\"text\"]\n",
    "input_length = data[\"input_length\"]\n",
    "label_length = data[\"label_length\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "d5781e47-266a-4de4-8767-241add7ce8e6",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "d5781e47-266a-4de4-8767-241add7ce8e6",
    "outputId": "ab59c85f-4bd9-4d09-bc75-3f197f821629",
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "33"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(chars)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "e50c4d44",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "e50c4d44",
    "outputId": "79346cdf-0ad8-45c0-9878-1a5417384871",
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 256 256 16 33\n",
      "1 256 256 19 39\n",
      "2 256 512 25 51\n",
      "3 512 512 31 63\n",
      "4 512 512 37 75\n"
     ]
    }
   ],
   "source": [
    "model = QuartzNet(\n",
    "    R_repeat=0,\n",
    "    in_channels=64,\n",
    "    out_channels=len(chars) + 1,  # for blank token (ctc loss)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "bb80afc3-870d-4c02-ad26-0d87b6a9e9ef",
   "metadata": {
    "id": "bb80afc3-870d-4c02-ad26-0d87b6a9e9ef",
    "tags": []
   },
   "outputs": [],
   "source": [
    "# model = torch.load('quartznet_5x5.model')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "4c01c614",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "4c01c614",
    "outputId": "fd490b43-fabf-46c5-9a6d-142239df6f93",
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The model has 6,718,306 trainable parameters\n"
     ]
    }
   ],
   "source": [
    "print(f\"The model has {count_parameters(model):,} trainable parameters\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "112bd1fa-a83c-4e62-b75b-f85a5573b651",
   "metadata": {
    "id": "112bd1fa-a83c-4e62-b75b-f85a5573b651",
    "tags": []
   },
   "outputs": [],
   "source": [
    "def collate_fn(batch):\n",
    "    melspecs = [row[\"melspec\"] for row in batch]\n",
    "    texts = [row[\"text\"] for row in batch]\n",
    "    input_lengths = [row[\"input_length\"] for row in batch]\n",
    "    label_lengths = [row[\"label_length\"] for row in batch]\n",
    "\n",
    "    specs = [torch.transpose(spec, 0, 1) for spec in melspecs]\n",
    "    specs = pad_sequence(specs, batch_first=True)\n",
    "    specs = torch.transpose(specs, 1, 2)\n",
    "\n",
    "    labels = pad_sequence(texts, batch_first=True)\n",
    "    return (\n",
    "        specs,\n",
    "        labels,\n",
    "        torch.tensor(input_lengths),\n",
    "        torch.tensor(label_lengths),\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "fe79ca40-4b56-40e4-b30e-b82500c5a188",
   "metadata": {
    "id": "fe79ca40-4b56-40e4-b30e-b82500c5a188",
    "tags": []
   },
   "outputs": [],
   "source": [
    "train_loader = DataLoader(\n",
    "    train_dataset,\n",
    "    batch_size=32,\n",
    "    shuffle=True,\n",
    "    collate_fn=collate_fn,\n",
    "    num_workers=0,\n",
    "    pin_memory=True,\n",
    ")\n",
    "\n",
    "test_loader = DataLoader(\n",
    "    test_dataset,\n",
    "    batch_size=32,\n",
    "    shuffle=False,\n",
    "    collate_fn=collate_fn,\n",
    "    num_workers=0,\n",
    "    pin_memory=True,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "4443a1a7-8aa8-4df5-9b78-e89380461a84",
   "metadata": {
    "id": "4443a1a7-8aa8-4df5-9b78-e89380461a84",
    "tags": []
   },
   "outputs": [],
   "source": [
    "# sum_stat = 0\n",
    "# for i in tqdm(train_dataset):\n",
    "#     sum_stat = i['melspec'].sum(1)\n",
    "\n",
    "# mean = sum_stat / len(train_dataset)\n",
    "\n",
    "# sum_stat = 0\n",
    "# for i in tqdm(train_dataset):\n",
    "#     sum_stat += ((i['melspec'] - mean.unsqueeze(1))**2).sum(1)\n",
    "\n",
    "# std = (sum_stat / len(train_dataset)) ** (0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "f08c519f-bc31-4b89-a931-96ce086ca1a8",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "f08c519f-bc31-4b89-a931-96ce086ca1a8",
    "outputId": "8379ac2f-3f39-4ce0-b32b-91337c3df72b",
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mel torch.Size([32, 64, 479])\n",
      "texts torch.Size([32, 64])\n",
      "torch.Size([32, 34, 240])\n"
     ]
    }
   ],
   "source": [
    "for batch in train_loader:\n",
    "    melspecs = batch[0]\n",
    "    texts = batch[1]\n",
    "    input_lengths = batch[2]\n",
    "    label_lengths = batch[3]\n",
    "    output = model(melspecs)\n",
    "    output = F.log_softmax(output, dim=1)\n",
    "    print(\"mel\", melspecs.shape)\n",
    "    print(\"texts\", texts.shape)\n",
    "    print(output.shape)\n",
    "    break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "bR-r77VCXWFu",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "bR-r77VCXWFu",
    "outputId": "234feb1e-c5a4-4e9e-dc65-977636d11d5a",
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor(0)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "melspecs.isinf().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "ded810fb-4f9c-4ce8-b401-08bd51f08593",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ded810fb-4f9c-4ce8-b401-08bd51f08593",
    "outputId": "7544056b-15c5-4e2e-d0a9-5102f560837c",
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([18,  6, 17,  9,  1, 12,  0, 15, 19,  6, 12, 29,  0,  3,  1,  3,  9, 12,\n",
       "        15, 14,  0, 18,  6,  8, 15, 14,  0, 25,  6, 18, 19, 29,  0, 24,  1, 18,\n",
       "        19, 29,  0, 25,  6, 18, 19, 14,  1,  5, 23,  1, 19, 29,  0,  0,  0,  0,\n",
       "         0,  0,  0,  0,  0,  0,  0,  0,  0,  0])"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "texts[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "9806f528-852b-4598-9edb-98f23bfde852",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 386
    },
    "id": "9806f528-852b-4598-9edb-98f23bfde852",
    "outputId": "75a30147-c2a3-49f7-d864-5b9e5f354db9",
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x1f8c9be66d0>"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 4))\n",
    "plt.imshow(melspecs[20].numpy(), aspect=\"auto\", origin=\"lower\", interpolation=\"none\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "eba20305",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "eba20305",
    "outputId": "1fec9a15-98e8-4fd9-e15e-c1e9eff135df",
    "tags": []
   },
   "outputs": [],
   "source": [
    "# output.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "0a1c483d",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "0a1c483d",
    "outputId": "fcc70ebc-d8f1-4adc-9a4a-32ecbc571ecb",
    "tags": []
   },
   "outputs": [],
   "source": [
    "# melspecs.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "2d0c559f",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "2d0c559f",
    "outputId": "5f06a92a-6d22-450e-f2a6-3226c0ed6821",
    "tags": []
   },
   "outputs": [],
   "source": [
    "# melspecs[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "731afb59",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "731afb59",
    "outputId": "0d4375c3-0ada-4549-d7ea-da8a4e5c46bd",
    "tags": []
   },
   "outputs": [],
   "source": [
    "# input_lengths"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "b4aedf3b-7d13-4765-b68a-9c83f82267de",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "b4aedf3b-7d13-4765-b68a-9c83f82267de",
    "outputId": "daaf16a5-c7b1-428a-d473-2d5fc7cb987f",
    "tags": []
   },
   "outputs": [],
   "source": [
    "# criterion = nn.CTCLoss(blank=len(chars), zero_infinity=True)\n",
    "# criterion(\n",
    "#     output.permute(2, 0, 1),\n",
    "#     texts,\n",
    "#     torch.ceil(input_lengths.float() / 2).int(),\n",
    "#     label_lengths,\n",
    "# )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "58a53bff-8a29-4649-b8e3-03170b70277f",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "58a53bff-8a29-4649-b8e3-03170b70277f",
    "outputId": "6d1c11c7-0b66-43d0-864a-8a3441f7f0d3",
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.cuda.is_available()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "7829a9aa-914a-497a-9eb1-3f25566b974d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2024-04-05 22:11:48,698 - root - INFO - cuda\n"
     ]
    }
   ],
   "source": [
    "device = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
    "logging.info(device)\n",
    "\n",
    "model = model.to(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "d7b02933-3cd0-4738-ae9f-58e1f701ed4b",
   "metadata": {
    "id": "d7b02933-3cd0-4738-ae9f-58e1f701ed4b",
    "tags": []
   },
   "outputs": [],
   "source": [
    "def train(\n",
    "    model: nn.Module,\n",
    "    train_loader: DataLoader,\n",
    "    criterion: nn.Module,\n",
    "    optimizer: torch.optim.Optimizer,\n",
    "    scheduler,\n",
    "    device: str,\n",
    "    clip: float = 3.0,\n",
    "    train_history=None,\n",
    "    valid_history=None,\n",
    "):\n",
    "    epoch_loss = 0\n",
    "    history = []\n",
    "\n",
    "    model.train()\n",
    "    for i, batch in enumerate(tqdm(train_loader)):\n",
    "        try:\n",
    "            melspecs = batch[0].to(device)\n",
    "            texts = batch[1].to(device)\n",
    "            input_lengths = batch[2].to(device)\n",
    "            label_lengths = batch[3].to(device)\n",
    "    \n",
    "            output = model(melspecs).permute(2, 0, 1)\n",
    "            output = F.log_softmax(output, dim=2)\n",
    "    \n",
    "            loss = criterion(\n",
    "                output,\n",
    "                texts,\n",
    "                torch.ceil(input_lengths.float() / 2).int(),\n",
    "                label_lengths,\n",
    "            )\n",
    "        except Exception as e:\n",
    "            print(e)\n",
    "            continue\n",
    "        epoch_loss += loss.item()\n",
    "\n",
    "        optimizer.zero_grad()\n",
    "        loss.backward()\n",
    "        # torch.nn.utils.clip_grad_norm_(model.parameters(), clip)\n",
    "        optimizer.step()\n",
    "        # scheduler.step()\n",
    "\n",
    "        history.append(loss.cpu().data.numpy())\n",
    "        if ((i + 1) % 30 == 0) or (i == (len(train_loader) - 1)):\n",
    "            fig, ax = plt.subplots(nrows=1, ncols=2, figsize=(12, 6))\n",
    "\n",
    "            clear_output(True)\n",
    "            ax[0].plot(history, label=\"train loss\")\n",
    "            ax[0].set_xlabel(\"Batch\")\n",
    "            ax[0].set_title(\"Train loss\")\n",
    "            if train_history is not None:\n",
    "                ax[1].plot(train_history, label=\"general train history\")\n",
    "                ax[1].set_xlabel(\"Epoch\")\n",
    "            if valid_history is not None:\n",
    "                ax[1].plot(valid_history, label=\"general valid history\")\n",
    "            plt.legend()\n",
    "\n",
    "            plt.show()\n",
    "        i += 1\n",
    "\n",
    "    return epoch_loss / len(train_loader), loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "26ce2190-997c-429a-893a-dbcddafc3437",
   "metadata": {
    "id": "26ce2190-997c-429a-893a-dbcddafc3437",
    "tags": []
   },
   "outputs": [],
   "source": [
    "def validate(\n",
    "    model: nn.Module, valid_loader: DataLoader, criterion: nn.Module, device: str\n",
    "):\n",
    "\n",
    "    epoch_loss = 0\n",
    "    model.eval()\n",
    "    with torch.no_grad():\n",
    "        for batch in tqdm(valid_loader):\n",
    "            melspecs = batch[0].to(device)\n",
    "            texts = batch[1].to(device)\n",
    "            input_lengths = batch[2].to(device)\n",
    "            label_lengths = batch[3].to(device)\n",
    "\n",
    "            output = model(melspecs).permute(2, 0, 1)\n",
    "            output = F.log_softmax(output, dim=2)\n",
    "            loss = criterion(\n",
    "                output,\n",
    "                texts,\n",
    "                torch.ceil(input_lengths.float() / 2).int(),\n",
    "                label_lengths,\n",
    "            )\n",
    "            epoch_loss += loss.item()\n",
    "\n",
    "    return epoch_loss / len(valid_loader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "31243c91-7b55-4dd1-80bd-81b116f23a9d",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def test(\n",
    "    model: nn.Module, dataset: Dataset, ind: int\n",
    "):\n",
    "    model.eval()\n",
    "    \n",
    "    with torch.no_grad():\n",
    "    \n",
    "        melspecs = dataset[ind][\"melspec\"]\n",
    "        texts = dataset[ind][\"text\"]\n",
    "        input_lengths = dataset[ind][\"input_length\"]\n",
    "        label_lengths = dataset[ind][\"label_length\"]\n",
    "\n",
    "        output = model(melspecs.to('cuda').unsqueeze(0)).permute(2, 0, 1)\n",
    "        output = F.log_softmax(output, dim=2)\n",
    "        output = output.permute(1, 0, 2)\n",
    "        output.shape\n",
    "\n",
    "        ground_true = ''.join([int_to_char[i] for i in texts.cpu().numpy().tolist()])\n",
    "        predicted = ''.join([int_to_char.get(i, '') for i in output.argmax(dim=2).cpu().numpy().tolist()[0]])\n",
    "\n",
    "        print(f'ground_true: {ground_true}')\n",
    "        print(f'predicted: {predicted}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "fa42e34d-8a34-4b06-b07e-51f453c0056f",
   "metadata": {
    "id": "fa42e34d-8a34-4b06-b07e-51f453c0056f",
    "tags": []
   },
   "outputs": [],
   "source": [
    "def fit(\n",
    "    model: nn.Module,\n",
    "    train_loader: DataLoader,\n",
    "    valid_loader: DataLoader,\n",
    "    criterion,\n",
    "    optimizer,\n",
    "    scheduler,\n",
    "    start_epoch: int,\n",
    "    end_epoch: int,\n",
    "    clip: float = 3.0,\n",
    "):\n",
    "    \n",
    "    train_losses = []\n",
    "    valid_losses = []\n",
    "\n",
    "    best_valid_loss = 0.9\n",
    "\n",
    "    for epoch in range(start_epoch, end_epoch):\n",
    "        train_loss, last_batch_loss = train(\n",
    "            model,\n",
    "            train_loader,\n",
    "            criterion,\n",
    "            optimizer,\n",
    "            scheduler,\n",
    "            device,\n",
    "            clip,\n",
    "            train_losses,\n",
    "            valid_losses,\n",
    "        )\n",
    "        valid_loss = validate(model, valid_loader, criterion, device)\n",
    "        logging.info(\n",
    "            f\"Epoch: {epoch}, Training Loss: {train_loss}, Valid Loss: {valid_loss}\"\n",
    "        )\n",
    "        \n",
    "        test(model, test_dataset, 300)\n",
    "\n",
    "        train_losses.append(train_loss)\n",
    "        valid_losses.append(valid_loss)\n",
    "        \n",
    "        if valid_loss < best_valid_loss:\n",
    "            torch.save(model, f\"quartznet_5x5_{epoch}.model\")\n",
    "            best_valid_loss = valid_loss\n",
    "\n",
    "        save_checkpoint(\n",
    "            last_epoch=epoch,\n",
    "            last_batch_idx=epoch * len(train_loader) - 1,\n",
    "            model=model,\n",
    "            optimizer=optimizer,\n",
    "            scheduler=scheduler,\n",
    "            loss=last_batch_loss,\n",
    "        )\n",
    "        logging.info('checkpoint created')\n",
    "        \n",
    "    return train_losses, valid_losses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "de0d7394-4748-4722-898f-b91a9c1c272a",
   "metadata": {},
   "outputs": [],
   "source": [
    "def save_checkpoint(\n",
    "    last_epoch: int,\n",
    "    last_batch_idx: int,\n",
    "    model,\n",
    "    optimizer,\n",
    "    scheduler,\n",
    "    loss,\n",
    "    path: str = \"quartznet\",\n",
    "    postfix: str = \"last_667\",\n",
    "):\n",
    "\n",
    "    if not os.path.exists(path):\n",
    "        os.mkdir(path)\n",
    "\n",
    "    torch.save({\n",
    "        'last_epoch': last_epoch,\n",
    "        'last_batch_idx': last_batch_idx,\n",
    "        'model_state_dict': model.state_dict(),\n",
    "        'optimizer_state_dict': optimizer.state_dict(),\n",
    "        'scheduler_state_dict': scheduler.state_dict(),\n",
    "        'loss': loss,\n",
    "        }, os.path.join(path, f\"model_{postfix}.checkpoint\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "264ec03a-403f-4b94-b1b0-4e92acca5c37",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_checkpoint(\n",
    "    model,\n",
    "    optimizer,\n",
    "    scheduler,\n",
    "    path: str = \"quartznet\",\n",
    "    postfix: str = \"last_667\",\n",
    "):\n",
    "    checkpoint = torch.load(os.path.join(path, f\"model_{postfix}.checkpoint\"))\n",
    "    \n",
    "    model.load_state_dict(checkpoint['model_state_dict'])\n",
    "    optimizer.load_state_dict(checkpoint['optimizer_state_dict'])\n",
    "    scheduler.load_state_dict(checkpoint['scheduler_state_dict'])\n",
    "    last_epoch = checkpoint['last_epoch']\n",
    "    last_batch_idx = checkpoint['last_batch_idx']\n",
    "    loss = checkpoint['loss']\n",
    "\n",
    "    return (\n",
    "        model,\n",
    "        optimizer,\n",
    "        scheduler,\n",
    "        last_epoch,\n",
    "        last_batch_idx,\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "97ce5d8a-ed0e-4b14-81ae-3dc7a8cb9510",
   "metadata": {},
   "outputs": [],
   "source": [
    "from_checkpoint = True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "8e780693-2f1c-4b5a-8e1c-dc660fb0bb76",
   "metadata": {},
   "outputs": [],
   "source": [
    "end_epoch = 10\n",
    "criterion = nn.CTCLoss(blank=len(chars), zero_infinity=True)\n",
    "if from_checkpoint:\n",
    "    optimizer = torch.optim.Adam(model.parameters(), lr=5e-4, weight_decay=0.0001)\n",
    "    scheduler = torch.optim.lr_scheduler.OneCycleLR(\n",
    "        optimizer,\n",
    "        max_lr=1e-3,\n",
    "        steps_per_epoch=len(train_loader),\n",
    "        epochs=end_epoch,\n",
    "        div_factor=25.0,\n",
    "        pct_start=0.3,\n",
    "        last_epoch=-1,\n",
    "    )\n",
    "    model, optimizer, scheduler, last_epoch, last_batch_idx = load_checkpoint(\n",
    "        model,\n",
    "        optimizer,\n",
    "        scheduler,\n",
    "    )\n",
    "    start_epoch = last_epoch + 1\n",
    "    scheduler.last_epoch = last_batch_idx\n",
    "else:\n",
    "    start_epoch = 0\n",
    "    optimizer = torch.optim.Adam(model.parameters(), lr=5e-4, weight_decay=0.0001)\n",
    "    scheduler = torch.optim.lr_scheduler.OneCycleLR(\n",
    "        optimizer,\n",
    "        max_lr=1e-3,\n",
    "        steps_per_epoch=len(train_loader),\n",
    "        epochs=end_epoch,\n",
    "        div_factor=25.0,\n",
    "        pct_start=0.3,\n",
    "        last_epoch=-1,\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "7ef441c2-7eeb-407e-8fa4-9be4220dde28",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "start_epoch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "e601cad0-d00b-4cd0-b763-8d76c12734a2",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 582
    },
    "id": "e601cad0-d00b-4cd0-b763-8d76c12734a2",
    "outputId": "6a0c4748-1b89-4507-8256-70c25c48744b",
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████| 12500/12500 [2:46:58<00:00,  1.25it/s]\n",
      "100%|████████████████████████████████████████████████████████████████████████████████| 372/372 [01:46<00:00,  3.51it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2024-04-06 01:03:21,027 - root - INFO - Epoch: 9, Training Loss: 0.9845657639980316, Valid Loss: 0.8699753078561957\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ground_true: можешь включить сорок девять дней\n",
      "predicted: можжеешшьь  вкклюючиитьь ссергени\n",
      "2024-04-06 01:03:21,298 - root - INFO - checkpoint created\n"
     ]
    }
   ],
   "source": [
    "train_losses, valid_losses = fit(\n",
    "    model=model,\n",
    "    train_loader=train_loader,\n",
    "    valid_loader=test_loader,\n",
    "    criterion=criterion,\n",
    "    optimizer=optimizer,\n",
    "    scheduler=scheduler,\n",
    "    start_epoch=start_epoch,\n",
    "    end_epoch=end_epoch,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "6603b58b-11e5-4961-a7fb-cd8bcbab686c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "QuartzNet(\n",
       "  (C1): Sequential(\n",
       "    (0): Conv1d(64, 64, kernel_size=(33,), stride=(2,), padding=(16,), groups=64, bias=False)\n",
       "    (1): Conv1d(64, 256, kernel_size=(1,), stride=(1,), bias=False)\n",
       "    (2): BatchNorm1d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "    (3): ReLU()\n",
       "    (4): Dropout(p=0.2, inplace=False)\n",
       "  )\n",
       "  (B): ModuleList(\n",
       "    (0): JasperBlock(\n",
       "      (blocks): Sequential(\n",
       "        (0): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(256, 256, kernel_size=(33,), stride=(1,), padding=(16,), groups=256, bias=False)\n",
       "            (1): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (1): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(256, 256, kernel_size=(33,), stride=(1,), padding=(16,), groups=256, bias=False)\n",
       "            (1): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (2): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(256, 256, kernel_size=(33,), stride=(1,), padding=(16,), groups=256, bias=False)\n",
       "            (1): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (3): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(256, 256, kernel_size=(33,), stride=(1,), padding=(16,), groups=256, bias=False)\n",
       "            (1): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (4): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(256, 256, kernel_size=(33,), stride=(1,), padding=(16,), groups=256, bias=False)\n",
       "            (1): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "      )\n",
       "      (residual): Sequential(\n",
       "        (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False)\n",
       "        (1): BatchNorm1d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (relu): ReLU()\n",
       "      (dropout): Dropout(p=0.2, inplace=False)\n",
       "    )\n",
       "    (1): JasperBlock(\n",
       "      (blocks): Sequential(\n",
       "        (0): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(256, 256, kernel_size=(39,), stride=(1,), padding=(19,), groups=256, bias=False)\n",
       "            (1): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (1): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(256, 256, kernel_size=(39,), stride=(1,), padding=(19,), groups=256, bias=False)\n",
       "            (1): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (2): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(256, 256, kernel_size=(39,), stride=(1,), padding=(19,), groups=256, bias=False)\n",
       "            (1): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (3): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(256, 256, kernel_size=(39,), stride=(1,), padding=(19,), groups=256, bias=False)\n",
       "            (1): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (4): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(256, 256, kernel_size=(39,), stride=(1,), padding=(19,), groups=256, bias=False)\n",
       "            (1): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "      )\n",
       "      (residual): Sequential(\n",
       "        (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False)\n",
       "        (1): BatchNorm1d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (relu): ReLU()\n",
       "      (dropout): Dropout(p=0.2, inplace=False)\n",
       "    )\n",
       "    (2): JasperBlock(\n",
       "      (blocks): Sequential(\n",
       "        (0): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(256, 256, kernel_size=(51,), stride=(1,), padding=(25,), groups=256, bias=False)\n",
       "            (1): Conv1d(256, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (1): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(512, 512, kernel_size=(51,), stride=(1,), padding=(25,), groups=512, bias=False)\n",
       "            (1): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (2): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(512, 512, kernel_size=(51,), stride=(1,), padding=(25,), groups=512, bias=False)\n",
       "            (1): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (3): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(512, 512, kernel_size=(51,), stride=(1,), padding=(25,), groups=512, bias=False)\n",
       "            (1): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (4): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(512, 512, kernel_size=(51,), stride=(1,), padding=(25,), groups=512, bias=False)\n",
       "            (1): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "      )\n",
       "      (residual): Sequential(\n",
       "        (0): Conv1d(256, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "        (1): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (relu): ReLU()\n",
       "      (dropout): Dropout(p=0.2, inplace=False)\n",
       "    )\n",
       "    (3): JasperBlock(\n",
       "      (blocks): Sequential(\n",
       "        (0): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(512, 512, kernel_size=(63,), stride=(1,), padding=(31,), groups=512, bias=False)\n",
       "            (1): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (1): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(512, 512, kernel_size=(63,), stride=(1,), padding=(31,), groups=512, bias=False)\n",
       "            (1): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (2): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(512, 512, kernel_size=(63,), stride=(1,), padding=(31,), groups=512, bias=False)\n",
       "            (1): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (3): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(512, 512, kernel_size=(63,), stride=(1,), padding=(31,), groups=512, bias=False)\n",
       "            (1): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (4): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(512, 512, kernel_size=(63,), stride=(1,), padding=(31,), groups=512, bias=False)\n",
       "            (1): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "      )\n",
       "      (residual): Sequential(\n",
       "        (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "        (1): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (relu): ReLU()\n",
       "      (dropout): Dropout(p=0.2, inplace=False)\n",
       "    )\n",
       "    (4): JasperBlock(\n",
       "      (blocks): Sequential(\n",
       "        (0): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(512, 512, kernel_size=(75,), stride=(1,), padding=(37,), groups=512, bias=False)\n",
       "            (1): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (1): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(512, 512, kernel_size=(75,), stride=(1,), padding=(37,), groups=512, bias=False)\n",
       "            (1): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (2): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(512, 512, kernel_size=(75,), stride=(1,), padding=(37,), groups=512, bias=False)\n",
       "            (1): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (3): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(512, 512, kernel_size=(75,), stride=(1,), padding=(37,), groups=512, bias=False)\n",
       "            (1): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "        (4): ConvBlock(\n",
       "          (layers): Sequential(\n",
       "            (0): Conv1d(512, 512, kernel_size=(75,), stride=(1,), padding=(37,), groups=512, bias=False)\n",
       "            (1): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "            (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (3): ReLU()\n",
       "            (4): Dropout(p=0.2, inplace=False)\n",
       "          )\n",
       "        )\n",
       "      )\n",
       "      (residual): Sequential(\n",
       "        (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "        (1): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (relu): ReLU()\n",
       "      (dropout): Dropout(p=0.2, inplace=False)\n",
       "    )\n",
       "  )\n",
       "  (C2): Sequential(\n",
       "    (0): Conv1d(512, 512, kernel_size=(87,), stride=(1,), padding=(86,), dilation=(2,), groups=512, bias=False)\n",
       "    (1): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False)\n",
       "    (2): BatchNorm1d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "    (3): ReLU()\n",
       "    (4): Dropout(p=0.2, inplace=False)\n",
       "  )\n",
       "  (C3): Sequential(\n",
       "    (0): Conv1d(512, 1024, kernel_size=(1,), stride=(1,), bias=False)\n",
       "    (1): BatchNorm1d(1024, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)\n",
       "    (2): ReLU()\n",
       "    (3): Dropout(p=0.2, inplace=False)\n",
       "  )\n",
       "  (C4): Conv1d(1024, 34, kernel_size=(1,), stride=(1,))\n",
       ")"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.eval()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "cd55ba01-e983-43b7-b984-33f3b997eeef",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 144
    },
    "id": "cd55ba01-e983-43b7-b984-33f3b997eeef",
    "outputId": "8b217526-0f10-4409-eab6-e0a91d68e685",
    "tags": []
   },
   "outputs": [],
   "source": [
    "torch.save(model, \"quartznet/quartznet_5x5_667.model\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "113115ae-ab09-4052-af12-a5b2b5a313b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "checkpoint = torch.load(os.path.join(\"quartznet\", f\"model_last_667.checkpoint\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "fe5ec454-df6c-4ceb-82bc-e371974a6189",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<All keys matched successfully>"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.load_state_dict(checkpoint['model_state_dict'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "656348e9-3511-4461-8d59-1b41cb71128a",
   "metadata": {
    "id": "656348e9-3511-4461-8d59-1b41cb71128a"
   },
   "outputs": [],
   "source": [
    "ind = 3\n",
    "melspecs = test_dataset[ind][\"melspec\"]\n",
    "texts = test_dataset[ind][\"text\"]\n",
    "input_lengths = test_dataset[ind][\"input_length\"]\n",
    "label_lengths = test_dataset[ind][\"label_length\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "d362a3f2-5f33-47c4-bd97-feeb2d292dc3",
   "metadata": {
    "id": "d362a3f2-5f33-47c4-bd97-feeb2d292dc3",
    "tags": []
   },
   "outputs": [],
   "source": [
    "ind = 30\n",
    "melspecs = train_dataset[ind][\"melspec\"]\n",
    "texts = train_dataset[ind][\"text\"]\n",
    "input_lengths = train_dataset[ind][\"input_length\"]\n",
    "label_lengths = train_dataset[ind][\"label_length\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "f5ab4858-69c5-427f-bdbd-23bbbec1199f",
   "metadata": {
    "id": "f5ab4858-69c5-427f-bdbd-23bbbec1199f",
    "outputId": "b831a3f9-2ff1-4890-c532-f42081789d58",
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([1, 146, 34])"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "output = model(melspecs.to(device).unsqueeze(0)).permute(2, 0, 1)\n",
    "output = F.log_softmax(output, dim=2)\n",
    "output = output.permute(1, 0, 2)\n",
    "output.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "1090e1bc-37ce-4074-a49c-7cc165d0a10c",
   "metadata": {
    "id": "1090e1bc-37ce-4074-a49c-7cc165d0a10c",
    "outputId": "a9f61018-2e78-4ee9-a4eb-f566890359ce",
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'марина федоровна'"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\".join([int_to_char[i] for i in texts.cpu().numpy().tolist()])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "1de50f56-f115-4d68-9991-0e5d4b8a0902",
   "metadata": {
    "id": "1de50f56-f115-4d68-9991-0e5d4b8a0902",
    "outputId": "5e6c3c7d-f7ec-46f1-e8b3-4b245a1eb9c2",
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x1f8d32aa4d0>"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 4))\n",
    "plt.imshow(melspecs.numpy(), aspect=\"auto\", origin=\"lower\", interpolation=\"none\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "98b544ac-7230-4b50-b0e3-25ce75a1a3e0",
   "metadata": {
    "id": "98b544ac-7230-4b50-b0e3-25ce75a1a3e0",
    "outputId": "b433e537-9c7a-4f1e-cc9f-4676fbb5005b",
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ммаа$$$рр$и$$$нн$а$$$$ $$$с$$е$$$$$д$оо$рооовввннна$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$'"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\".join(\n",
    "    [int_to_char.get(i, \"$\") for i in output.argmax(dim=2).cpu().numpy().tolist()[0]]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "1a2f17ce-d998-407d-b56a-18d243d21289",
   "metadata": {
    "id": "98b544ac-7230-4b50-b0e3-25ce75a1a3e0",
    "outputId": "b433e537-9c7a-4f1e-cc9f-4676fbb5005b",
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'ммаарринна седоорооовввннна'"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\".join(\n",
    "    [int_to_char.get(i, \"\") for i in output.argmax(dim=2).cpu().numpy().tolist()[0]]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "34ef3809-c5f3-4c66-844d-8a2d0d92039e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2024-04-06 01:04:39,348 - root - INFO - 6813: курс тысяч казахских тенге к английскому фунту на сегодня\n",
      "ккурррсь  ттыыссяячччи  поззаххсскиххх   пеннге каннлийскомму  фундув  на ссеггодддння\n",
      "2024-04-06 01:04:39,445 - root - INFO - 2947: у тебя в запасе есть спасти санту\n",
      "у  ттееббяя вв  ззаппассее    есстттьь   пасстисаант\n",
      "2024-04-06 01:04:39,496 - root - INFO - 2899: фэнтези\n",
      "файктззии\n",
      "2024-04-06 01:04:39,561 - root - INFO - 8494: киношка люди в черном три покажи\n",
      "кииноошка  любивчоррнном трри  ппокажиии\n",
      "2024-04-06 01:04:39,616 - root - INFO - 2685: покажи мне встречу апл манчестер против манчестер сити\n",
      "пппокажи  ммнее всстрреччу  апплллл   ммааннчеесссттеерр  ппрооттиввв мманннччеесссттеерр сситии\n",
      "2024-04-06 01:04:39,678 - root - INFO - 2099: музыка ютуб\n",
      "ммозыкаа ютубб\n",
      "2024-04-06 01:04:39,736 - root - INFO - 1518: джой запиши меня к терапевту на понедельник\n",
      "ддджжоййй   зааппееше меення ктееррпетта на  понденик\n",
      "2024-04-06 01:04:39,789 - root - INFO - 9437: севак ханагян на ютюбе\n",
      "фивоокмаге  ннаа   ююютттюббеее\n",
      "2024-04-06 01:04:39,846 - root - INFO - 6363: сезон один сериала сабрина маленькая ведьма\n",
      "ссееззоонннн  ооддиинннн  ссеерррииаааалллллаа   с апрриноо  ммаленнкая ветьва\n",
      "2024-04-06 01:04:39,907 - root - INFO - 820: люди как мы можешь включить\n",
      "ллудикауткман можешьь ввкллююччиитььь\n"
     ]
    }
   ],
   "source": [
    "model.eval()\n",
    "for ind in np.random.randint(0, len(test_dataset), size=10):\n",
    "    melspecs = test_dataset[ind][\"melspec\"]\n",
    "    texts = test_dataset[ind][\"text\"]\n",
    "    input_lengths = test_dataset[ind][\"input_length\"]\n",
    "    label_lengths = test_dataset[ind][\"label_length\"]\n",
    "\n",
    "    output = model(melspecs.unsqueeze(0).to(device)).permute(2, 0, 1)\n",
    "    output = F.log_softmax(output, dim=2)\n",
    "    output = output.permute(1, 0, 2)\n",
    "\n",
    "    original_text = \"\".join([int_to_char[i] for i in texts.cpu().numpy().tolist()])\n",
    "\n",
    "    predicted_text = \"\".join(\n",
    "        [int_to_char.get(i, \"\") for i in output.argmax(dim=2).cpu().numpy().tolist()[0]]\n",
    "    )\n",
    "\n",
    "    logging.info(f'{ind}: {original_text}\\n{predicted_text}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "0572cecc-341a-4f55-90ee-a5da1e8bd865",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2024-04-06 01:06:46,237 - root - INFO - 1074: где находится девять акров\n",
      "ггдде  нахходдииттссяя  деввячкка\n",
      "2024-04-06 01:06:46,290 - root - INFO - 4670: включи биг бэнг\n",
      "ввкклллюччии  бегданк\n",
      "2024-04-06 01:06:46,343 - root - INFO - 3550: хочу на телевизоре канал амедиа один\n",
      "чуу ннаа   ттеелллеевввииззоорррее  ккканнааллл  амеегииодиинн\n",
      "2024-04-06 01:06:46,391 - root - INFO - 9637: вилюйске\n",
      "ввилевойсске\n",
      "2024-04-06 01:06:46,424 - root - INFO - 9192: поиск любимые команды манчестер юнайтед матчи\n",
      "ожжа крлюббиемный  ккооммаада   мманннччеееттерр    юююннайййттеедд   маати\n",
      "2024-04-06 01:06:46,477 - root - INFO - 2019: джой долгота измайловский парк\n",
      "дджжжооййй  далнаа  изнаасипа\n",
      "2024-04-06 01:06:46,537 - root - INFO - 4549: давайте втроем\n",
      "довааййки  тровеем\n",
      "2024-04-06 01:06:46,586 - root - INFO - 607: найти лучшие фильмы жанра драма\n",
      "ннайййттии    ллллуучччшшшииее  фффиилльььммыыы    жжаанннрррааа   дррааммаа\n",
      "2024-04-06 01:06:46,632 - root - INFO - 11494: сбер следующая\n",
      "ссбберрр  ссеб\n",
      "2024-04-06 01:06:46,679 - root - INFO - 6469: выведи мне падение последней империи ультра эйч ди\n",
      "ввыывведии  ммнннее  подеьнни  послледдни  перриутрроч\n"
     ]
    }
   ],
   "source": [
    "model.eval()\n",
    "for ind in np.random.randint(0, len(test_dataset), size=10):\n",
    "    melspecs = test_dataset[ind][\"melspec\"]\n",
    "    texts = test_dataset[ind][\"text\"]\n",
    "    input_lengths = test_dataset[ind][\"input_length\"]\n",
    "    label_lengths = test_dataset[ind][\"label_length\"]\n",
    "\n",
    "    output = model(melspecs.unsqueeze(0).to(device)).permute(2, 0, 1)\n",
    "    output = F.log_softmax(output, dim=2)\n",
    "    output = output.permute(1, 0, 2)\n",
    "\n",
    "    original_text = \"\".join([int_to_char[i] for i in texts.cpu().numpy().tolist()])\n",
    "\n",
    "    predicted_text = \"\".join(\n",
    "        [int_to_char.get(i, \"\") for i in output.argmax(dim=2).cpu().numpy().tolist()[0]]\n",
    "    )\n",
    "\n",
    "    logging.info(f'{ind}: {original_text}\\n{predicted_text}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d35273c3-ebcf-4e75-b434-61e70e9e6c48",
   "metadata": {
    "id": "d35273c3-ebcf-4e75-b434-61e70e9e6c48"
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}