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/voc3
TianYuan 2b01e40525
[TTS]soft link for shell in example, add skip_copy_wave in norm stage of G… (#2851)
2 years ago
..
conf add ernie sat model file and config 2 years ago
local [TTS]soft link for shell in example, add skip_copy_wave in norm stage of G… (#2851) 2 years ago
README.md [TTS]Add TTS Paddle-Lite x86 inference (#2667) 2 years ago
finetune.sh [TTS]soft link for shell in example, add skip_copy_wave in norm stage of G… (#2851) 2 years ago
path.sh merge deepspeech, parakeet and text_processing into paddlespeech 3 years ago
run.sh [TTS]Add slim for TTS (#2729) 2 years ago

README.md

Multi Band MelGAN with CSMSC

This example contains code used to train a Multi Band MelGAN 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 results to cut the silence in the edge of audio. You can download from here baker_alignment_tone.tar.gz, or train your MFA model reference to mfa example of our repo.

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.
./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
├── test
│   ├── norm
│   └── raw
└── train
    ├── norm
    ├── raw
    └── feats_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 the log magnitude of the mel spectrogram of each utterance, while the norm folder contains the normalized spectrogram. The statistics used to normalize the spectrogram are computed from the training set, which is located in dump/train/feats_stats.npy.

Also, there is a metadata.jsonl in each subfolder. It is a table-like file that contains id and paths to the spectrogram of each utterance.

Model Training

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

./local/train.sh calls ${BIN_DIR}/train.py. Here's the complete help message.

usage: train.py [-h] [--config CONFIG] [--train-metadata TRAIN_METADATA]
                [--dev-metadata DEV_METADATA] [--output-dir OUTPUT_DIR]
                [--ngpu NGPU]

Train a Multi-Band MelGAN model.

optional arguments:
  -h, --help            show this help message and exit
  --config CONFIG       Multi-Band MelGAN config file.
  --train-metadata TRAIN_METADATA
                        training data.
  --dev-metadata DEV_METADATA
                        dev data.
  --output-dir OUTPUT_DIR
                        output dir.
  --ngpu NGPU           if ngpu == 0, use cpu.
  1. --config is a config file in yaml format to overwrite the default config, which can be found at conf/default.yaml.
  2. --train-metadata and --dev-metadata should be the metadata file in the normalized subfolder of train and dev in the dump folder.
  3. --output-dir is the directory to save the results of the experiment. Checkpoints are saved in checkpoints/ inside this directory.
  4. --ngpu is the number of gpus to use, if ngpu == 0, use cpu.

Synthesizing

./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}
usage: synthesize.py [-h] [--generator-type GENERATOR_TYPE] [--config CONFIG]
                     [--checkpoint CHECKPOINT] [--test-metadata TEST_METADATA]
                     [--output-dir OUTPUT_DIR] [--ngpu NGPU]

Synthesize with GANVocoder.

optional arguments:
  -h, --help            show this help message and exit
  --generator-type GENERATOR_TYPE
                        type of GANVocoder, should in {pwgan, mb_melgan,
                        style_melgan, } now
  --config CONFIG       GANVocoder config file.
  --checkpoint CHECKPOINT
                        snapshot to load.
  --test-metadata TEST_METADATA
                        dev data.
  --output-dir OUTPUT_DIR
                        output dir.
  --ngpu NGPU           if ngpu == 0, use cpu.
  1. --config multi band melgan config file. You should use the same config with which the model is trained.
  2. --checkpoint is the checkpoint to load. Pick one of the checkpoints from checkpoints inside the training output directory.
  3. --test-metadata is the metadata of the test dataset. Use the metadata.jsonl in the dev/norm subfolder from the processed directory.
  4. --output-dir is the directory to save the synthesized audio files.
  5. --ngpu is the number of gpus to use, if ngpu == 0, use cpu.

Fine-tuning

Since there is no noise in the input of Multi-Band MelGAN, the audio quality is not so good (see espnet issue), we refer to the method proposed in HiFiGAN, finetune Multi-Band MelGAN with the predicted mel-spectrogram from FastSpeech2.

The length of mel-spectrograms should align with the length of wavs, so we should generate mels using ground truth alignment.

But since we are fine-tuning, we should use the statistics computed during the training step.

You should first download pretrained FastSpeech2 model from fastspeech2_nosil_baker_ckpt_0.4.zip and unzip it.

Assume the path to the dump-dir of training step is dump. Assume the path to the duration result of CSMSC is durations.txt (generated during the training step's preprocessing). Assume the path to the pretrained FastSpeech2 model is fastspeech2_nosil_baker_ckpt_0.4.
The finetune.sh can

  1. source path.
  2. generate ground truth alignment mels.
  3. link *_wave.npy from dump to dump_finetune (because we only use new mels, the wavs are the ones used during the training step).
  4. copy features' stats from dump to dump_finetune.
  5. normalize the ground truth alignment mels.
  6. finetune the model.

Before finetune, make sure that the pretrained model is in finetune.sh 's ${output-dir}/checkpoints, and there is a records.jsonl in it to refer to this pretrained model

exp/finetune/checkpoints
├── records.jsonl
└── snapshot_iter_1000000.pdz

The content of records.jsonl should be as follows (change "path" to your ckpt path):

{"time": "2021-11-21 15:11:20.337311", "path": "~/PaddleSpeech/examples/csmsc/voc3/exp/finetune/checkpoints/snapshot_iter_1000000.pdz", "iteration": 1000000}

Run the command below

./finetune.sh

By default, finetune.sh will use conf/finetune.yaml as config, the dump-dir is dump_finetune, the experiment dir is exp/finetune.

TODO: The hyperparameter of finetune.yaml is not good enough, a smaller learning_rate should be used (more milestones should be set).

Pretrained Models

The pretrained model can be downloaded here:

The finetuned model can be downloaded here:

The static model can be downloaded here:

The ONNX model can be downloaded here:

The Paddle-Lite model can be downloaded here:

Model Step eval/generator_loss eval/log_stft_magnitude_loss eval/spectral_convergence_loss eval/sub_log_stft_magnitude_loss eval/sub_spectral_convergence_loss
default 1(gpu) x 1000000 2.4851 0.71778 0.2761 0.66334 0.2777
finetune 1(gpu) x 1000000 3.196967 0.977804 0.778484 0.889576 0.776756

Multi Band MelGAN checkpoint contains files listed below.

mb_melgan_csmsc_ckpt_0.1.1
├── default.yaml                  # default config used to train multi band melgan
├── feats_stats.npy               # statistics used to normalize spectrogram when training multi band melgan
└── snapshot_iter_1000000.pdz     # generator parameters of multi band melgan

Acknowledgement

We adapted some code from https://github.com/kan-bayashi/ParallelWaveGAN.