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/csmsc/tts3_rhy
HuangLiangJie a874d8f325
Add prosody prediction in synthesize_e2e, test=tts (#2693)
2 years ago
..
conf Add prosody prediction in synthesize_e2e, test=tts (#2693) 2 years ago
local Add prosody prediction in synthesize_e2e, test=tts (#2693) 2 years ago
README.md Add prosody prediction in synthesize_e2e, test=tts (#2693) 2 years ago
path.sh Add prosody prediction in synthesize_e2e, test=tts (#2693) 2 years ago
run.sh Add prosody prediction in synthesize_e2e, test=tts (#2693) 2 years ago

README.md

This example mainly follows the FastSpeech2 with CSMSC

This example contains code used to train a rhythm version of Fastspeech2 model with Chinese Standard Mandarin Speech Copus.

Dataset

Download and Extract

Download CSMSC from it's Official Website and extract it to ~/datasets. Then the dataset is in the directory ~/datasets/BZNSYP.

Get MFA Result and Extract

We use MFA to get durations for fastspeech2. You can directly download the rhythm version of MFA result from here baker_alignment_tone.zip, or train your MFA model reference to mfa example of our repo. Remember in our repo, you should add --rhy-with-duration flag to obtain the rhythm information.

Get Started

Assume the path to the dataset is ~/datasets/BZNSYP. Assume the path to the MFA result of CSMSC is ./baker_alignment_tone. Run the command below to

  1. source path.
  2. preprocess the dataset.
  3. train the model.
  4. synthesize wavs.
    • synthesize waveform from metadata.jsonl.
    • synthesize waveform from a text file.
  5. inference using the static model.
./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

./local/preprocess.sh ${conf_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
├── phone_id_map.txt
├── speaker_id_map.txt
├── test
│   ├── norm
│   └── raw
└── train
    ├── energy_stats.npy
    ├── norm
    ├── pitch_stats.npy
    ├── raw
    └── speech_stats.npy

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、pitch and energy 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, the path of pitch features, the path of energy features, speaker, and the id of each utterance.

For more details, You can refer to FastSpeech2 with CSMSC

Pretrained Model

Pretrained FastSpeech2 model for end-to-end rhythm version:

This FastSpeech2 checkpoint contains files listed below.

fastspeech2_rhy_csmsc_ckpt_1.3.0
├── default.yaml             # default config used to train fastspeech2
├── phone_id_map.txt         # phone vocabulary file when training fastspeech2
├── snapshot_iter_153000.pdz # model parameters and optimizer states
├── durations.txt            # the intermediate output of preprocess.sh
├── energy_stats.npy
├── pitch_stats.npy
└── speech_stats.npy         # statistics used to normalize spectrogram when training fastspeech2