OpenVoice/demo_part1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "b6ee1ede",
   "metadata": {},
   "source": [
    "## Voice Style Control Demo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "b7f043ee",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/data/zwl/anaconda3/envs/openvoice/lib/python3.9/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import torch\n",
    "import se_extractor\n",
    "from api import BaseSpeakerTTS, ToneColorConverter"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "15116b59",
   "metadata": {},
   "source": [
    "### Initialization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "aacad912",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loaded checkpoint 'checkpoints/base_speakers/EN/checkpoint.pth'\n",
      "missing/unexpected keys: [] []\n",
      "Loaded checkpoint 'checkpoints/converter/checkpoint.pth'\n",
      "missing/unexpected keys: [] []\n"
     ]
    }
   ],
   "source": [
    "ckpt_base = 'checkpoints/base_speakers/EN'\n",
    "ckpt_converter = 'checkpoints/converter'\n",
    "device = 'cuda:0'\n",
    "output_dir = 'outputs'\n",
    "\n",
    "base_speaker_tts = BaseSpeakerTTS(f'{ckpt_base}/config.json', device=device)\n",
    "base_speaker_tts.load_ckpt(f'{ckpt_base}/checkpoint.pth')\n",
    "\n",
    "tone_color_converter = ToneColorConverter(f'{ckpt_converter}/config.json', device=device)\n",
    "tone_color_converter.load_ckpt(f'{ckpt_converter}/checkpoint.pth')\n",
    "\n",
    "os.makedirs(output_dir, exist_ok=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7f67740c",
   "metadata": {},
   "source": [
    "### Obtain Tone Color Embedding"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f8add279",
   "metadata": {},
   "source": [
    "The `source_se` is the tone color embedding of the base speaker. \n",
    "It is an average of multiple sentences generated by the base speaker. We directly provide the result here but\n",
    "the readers feel free to extract `source_se` by themselves."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "63ff6273",
   "metadata": {},
   "outputs": [],
   "source": [
    "source_se = torch.load(f'{ckpt_base}/en_default_se.pth').to(device)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4f71fcc3",
   "metadata": {},
   "source": [
    "The `reference_speaker.mp3` below points to the short audio clip of the reference whose voice we want to clone. We provide an example here. If you use your own reference speakers, please **make sure each speaker has a unique filename.** The `se_extractor` will save the `targeted_se` using the filename of the audio and **will not automatically overwrite.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "55105eae",
   "metadata": {},
   "outputs": [],
   "source": [
    "reference_speaker = 'resources/example_reference.mp3'\n",
    "target_se, audio_name = se_extractor.get_se(reference_speaker, tone_color_converter)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a40284aa",
   "metadata": {},
   "source": [
    "### Inference"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "73dc1259",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " > Text splitted to sentences.\n",
      "This audio is generated by open voice.\n",
      " > ===========================\n",
      "ðɪs ˈɑdiˌoʊ ɪz ˈdʒɛnəɹˌeɪtɪd baɪ ˈoʊpən vɔɪs.\n",
      " length:45\n",
      " length:45\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/data/zwl/anaconda3/envs/openvoice/lib/python3.9/site-packages/wavmark/models/my_model.py:25: UserWarning: istft will require a complex-valued input tensor in a future PyTorch release. Matching the output from stft with return_complex=True.  (Triggered internally at /opt/conda/conda-bld/pytorch_1670525539683/work/aten/src/ATen/native/SpectralOps.cpp:978.)\n",
      "  return torch.istft(signal_wmd_fft, n_fft=self.n_fft, hop_length=self.hop_length, window=window,\n"
     ]
    }
   ],
   "source": [
    "save_path = f'{output_dir}/output_en_default.wav'\n",
    "\n",
    "# Run the base speaker tts\n",
    "text = \"This audio is generated by open voice.\"\n",
    "src_path = f'{output_dir}/tmp.wav'\n",
    "base_speaker_tts.tts(text, src_path, speaker='default', language='English', speed=1.0)\n",
    "\n",
    "# Run the tone color converter\n",
    "encode_message = \"@MyShell\"\n",
    "tone_color_converter.convert(\n",
    "    audio_src_path=src_path, \n",
    "    src_se=source_se, \n",
    "    tgt_se=target_se, \n",
    "    output_path=save_path,\n",
    "    message=encode_message)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6e3ea28a",
   "metadata": {},
   "source": [
    "**Try with different styles and speed.** The style can be controlled by the `speaker` parameter in the `base_speaker_tts.tts` method. Available choices: friendly, cheerful, excited, sad, angry, terrified, shouting, whispering. Note that the tone color embedding need to be updated. The speed can be controlled by the `speed` parameter. Let's try whispering with speed 0.9."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "fd022d38",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " > Text splitted to sentences.\n",
      "This audio is generated by open voice with a half-performance model.\n",
      " > ===========================\n",
      "ðɪs ˈɑdiˌoʊ ɪz ˈdʒɛnəɹˌeɪtɪd baɪ ˈoʊpən vɔɪs wɪθ ə half-peɹfoɹmance* ˈmɑdəɫ.\n",
      " length:76\n",
      " length:75\n"
     ]
    }
   ],
   "source": [
    "source_se = torch.load(f'{ckpt_base}/en_style_se.pth').to(device)\n",
    "save_path = f'{output_dir}/output_whispering.wav'\n",
    "\n",
    "# Run the base speaker tts\n",
    "text = \"This audio is generated by open voice with a half-performance model.\"\n",
    "src_path = f'{output_dir}/tmp.wav'\n",
    "base_speaker_tts.tts(text, src_path, speaker='whispering', language='English', speed=0.9)\n",
    "\n",
    "# Run the tone color converter\n",
    "encode_message = \"@MyShell\"\n",
    "tone_color_converter.convert(\n",
    "    audio_src_path=src_path, \n",
    "    src_se=source_se, \n",
    "    tgt_se=target_se, \n",
    "    output_path=save_path,\n",
    "    message=encode_message)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5fcfc70b",
   "metadata": {},
   "source": [
    "**Try with different languages.** OpenVoice can achieve multi-lingual voice cloning by simply replace the base speaker. We provide an example with a Chinese base speaker here and we encourage the readers to try `demo_part2.ipynb` for a detaied demo."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "a71d1387",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loaded checkpoint 'checkpoints/base_speakers/ZH/checkpoint.pth'\n",
      "missing/unexpected keys: [] []\n",
      " > Text splitted to sentences.\n",
      "今天天气真好, 我们一起出去吃饭吧.\n",
      " > ===========================\n",
      "tʃ⁼in→tʰjɛn→tʰjɛn→tʃʰi↓ ts`⁼ən→ xɑʊ↓↑,  wo↓↑mən i↓tʃʰi↓↑ ts`ʰu→tʃʰɥ↓ ts`ʰɹ`→fan↓ p⁼a.\n",
      " length:85\n",
      " length:85\n"
     ]
    }
   ],
   "source": [
    "\n",
    "ckpt_base = 'checkpoints/base_speakers/ZH'\n",
    "base_speaker_tts = BaseSpeakerTTS(f'{ckpt_base}/config.json', device=device)\n",
    "base_speaker_tts.load_ckpt(f'{ckpt_base}/checkpoint.pth')\n",
    "\n",
    "source_se = torch.load(f'{ckpt_base}/zh_default_se.pth').to(device)\n",
    "save_path = f'{output_dir}/output_chinese.wav'\n",
    "\n",
    "# Run the base speaker tts\n",
    "text = \"今天天气真好，我们一起出去吃饭吧。\"\n",
    "src_path = f'{output_dir}/tmp.wav'\n",
    "base_speaker_tts.tts(text, src_path, speaker='default', language='Chinese', speed=1.0)\n",
    "\n",
    "# Run the tone color converter\n",
    "encode_message = \"@MyShell\"\n",
    "tone_color_converter.convert(\n",
    "    audio_src_path=src_path, \n",
    "    src_se=source_se, \n",
    "    tgt_se=target_se, \n",
    "    output_path=save_path,\n",
    "    message=encode_message)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8e513094",
   "metadata": {},
   "source": [
    "**Tech for good.** For people who will deploy OpenVoice for public usage: We offer you the option to add watermark to avoid potential misuse. Please see the ToneColorConverter class. **MyShell reserves the ability to detect whether an audio is generated by OpenVoice**, no matter whether the watermark is added or not."
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "9d70c38e1c0b038dbdffdaa4f8bfa1f6767c43760905c87a9fbe7800d18c6c35"
  },
  "kernelspec": {
   "display_name": "Python 3.9.18 ('openvoice')",
   "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": "3.9.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}