You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
PaddleSpeech/examples/aishell3/vc0
TianYuan b5a7c2d080
update readme for aishell3_vc0, test=doc
3 years ago
..
conf update readme for aishell3_vc0, test=doc 3 years ago
local update tacotron2 voice cloning in aishell3 with new tacotron2, test=tts (#1419) 3 years ago
README.md update readme for aishell3_vc0, test=doc 3 years ago
path.sh rename tacotron2, test=tts 3 years ago
run.sh update tacotron2 voice cloning in aishell3 with new tacotron2, test=tts (#1419) 3 years ago

README.md

Tacotron2 + AISHELL-3 Voice Cloning

This example contains code used to train a Tacotron2 model with AISHELL-3. The trained model can be used in Voice Cloning Task, We refer to the model structure of Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis. The general steps are as follows:

  1. Speaker Encoder: We use Speaker Verification to train a speaker encoder. Datasets used in this task are different from those used in Tacotron2 because the transcriptions are not needed, we use more datasets, refer to ge2e.
  2. Synthesizer: We use the trained speaker encoder to generate speaker embedding for each sentence in AISHELL-3. This embedding is an extra input of Tacotron2 which will be concated with encoder outputs.
  3. Vocoder: We use Parallel Wave GAN as the neural Vocoder, refer to voc1.

Dataset

Download and Extract

Download AISHELL-3.

wget https://www.openslr.org/resources/93/data_aishell3.tgz

Extract AISHELL-3.

mkdir data_aishell3
tar zxvf data_aishell3.tgz -C data_aishell3

Get MFA Result and Extract

We use MFA2.x to get phonemes for Tacotron2, the durations of MFA are not needed here. You can download from here aishell3_alignment_tone.tar.gz, or train your MFA model reference to mfa example (use MFA1.x now) of our repo.

Pretrained GE2E Model

We use pretrained GE2E model to generate speaker embedding for each sentence.

Download pretrained GE2E model from here ge2e_ckpt_0.3.zip, and unzip it.

Get Started

Assume the path to the dataset is ~/datasets/data_aishell3. Assume the path to the MFA result of AISHELL-3 is ./aishell3_alignment_tone. Assume the path to the pretrained ge2e model is ./ge2e_ckpt_0.3.

Run the command below to

  1. source path.
  2. preprocess the dataset.
  3. train the model.
  4. synthesize waveform from metadata.jsonl.
  5. start a voice cloning inference.
./run.sh

You can choose a range of stages you want to run, or set stage equal to stop-stage to use only one stage, for example, running the following command will only preprocess the dataset.

./run.sh --stage 0 --stop-stage 0

Data Preprocessing

CUDA_VISIBLE_DEVICES=${gpus} ./local/preprocess.sh ${conf_path} ${ge2e_ckpt_path}

When it is done. A dump folder is created in the current directory. The structure of the dump folder is listed below.

dump
├── dev
│   ├── norm
│   └── raw
├── embed
│   ├── SSB0005
│   ├── SSB0009
│   ├── ...
│   └── ...
├── phone_id_map.txt
├── speaker_id_map.txt
├── test
│   ├── norm
│   └──  raw
└── train
    ├── norm
    ├── raw
    └── speech_stats.npy

The embed contains the generated speaker embedding for each sentence in AISHELL-3, which has the same file structure with wav files and the format is .npy.

The computing time of utterance embedding can be x hours.

The dataset is split into 3 parts, namely train, dev, and test, each of which contains a norm and raw subfolder. The raw folder contains speech features of each utterance, while the norm folder contains normalized ones. The statistics used to normalize features are computed from the training set, which is located in dump/train/*_stats.npy.

Also, there is a metadata.jsonl in each subfolder. It is a table-like file that contains phones, text_lengths, speech_lengths, durations, the path of speech features, speaker, and id of each utterance.

The preprocessing step is very similar to that one of tts0, but there is one more ge2e/inference step here.

Model Training

./local/train.sh calls ${BIN_DIR}/train.py.

CUDA_VISIBLE_DEVICES=${gpus} ./local/train.sh ${conf_path} ${train_output_path}

The training step is very similar to that one of tts0, but we should set --voice-cloning=True when calling ${BIN_DIR}/train.py.

Synthesizing

We use parallel wavegan as the neural vocoder. Download pretrained parallel wavegan model from pwg_aishell3_ckpt_0.5.zip and unzip it.

unzip pwg_aishell3_ckpt_0.5.zip

Parallel WaveGAN checkpoint contains files listed below.

pwg_aishell3_ckpt_0.5
├── default.yaml                   # default config used to train parallel wavegan
├── feats_stats.npy                # statistics used to normalize spectrogram when training parallel wavegan
└── snapshot_iter_1000000.pdz      # generator parameters of parallel wavegan

./local/synthesize.sh calls ${BIN_DIR}/../synthesize.py, which can synthesize waveform from metadata.jsonl.

CUDA_VISIBLE_DEVICES=${gpus} ./local/synthesize.sh ${conf_path} ${train_output_path} ${ckpt_name}

The synthesizing step is very similar to that one of tts0, but we should set --voice-cloning=True when calling ${BIN_DIR}/../synthesize.py.

Voice Cloning

Assume there are some reference audios in ./ref_audio

ref_audio
├── 001238.wav
├── LJ015-0254.wav
└── audio_self_test.mp3

./local/voice_cloning.sh calls ${BIN_DIR}/../voice_cloning.py

CUDA_VISIBLE_DEVICES=${gpus} ./local/voice_cloning.sh ${conf_path} ${train_output_path} ${ckpt_name} ${ge2e_params_path} ${ref_audio_dir}

Pretrained Model

tacotron2_aishell3_ckpt_vc0_0.2.0.zip

Model Step eval/loss eval/l1_loss eval/mse_loss eval/bce_loss eval/attn_loss
default 2(gpu) x 37596 0.58704 0.39623 0.15073 0.039 1.9981e-04

Tacotron2 checkpoint contains files listed below. (There is no need for speaker_id_map.txt here )

tacotron2_aishell3_ckpt_vc0_0.2.0
├── default.yaml            # default config used to train tacotron2
├── phone_id_map.txt        # phone vocabulary file when training tacotron2
├── snapshot_iter_37596.pdz # model parameters and optimizer states
└── speech_stats.npy        # statistics used to normalize spectrogram when training tacotron2

More

We strongly recommend that you use FastSpeech2 + AISHELL-3 Voice Cloning which works better.