msb-public
/
PaddleSpeech
mirror of https://github.com/PaddlePaddle/PaddleSpeech
1
0
Code Issues Projects Releases Wiki Activity

merge the develop

Browse Source
pull/1225/head
huangyuxin 3 years ago
parent 2c5902d7c5 42c109216d
commit fb6d1e2c11
26 changed files with 139 additions and 153 deletions
Show Stats Download Patch File Download Diff File

3
docs/source/released_model.md
Unescape View File

@ -49,7 +49,8 @@ Parallel WaveGAN| CSMSC |[PWGAN-csmsc](https://github.com/PaddlePaddle/PaddleSpe
Parallel WaveGAN| LJSpeech |[PWGAN-ljspeech](https://github.com/PaddlePaddle/PaddleSpeech/tree/develop/examples/ljspeech/voc1)|[pwg_ljspeech_ckpt_0.5.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/pwgan/pwg_ljspeech_ckpt_0.5.zip)|||
Parallel WaveGAN|AISHELL-3 |[PWGAN-aishell3](https://github.com/PaddlePaddle/PaddleSpeech/tree/develop/examples/aishell3/voc1)|[pwg_aishell3_ckpt_0.5.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/pwgan/pwg_aishell3_ckpt_0.5.zip)|||
Parallel WaveGAN| VCTK |[PWGAN-vctk](https://github.com/PaddlePaddle/PaddleSpeech/tree/develop/examples/vctk/voc1)|[pwg_vctk_ckpt_0.5.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/pwgan/pwg_vctk_ckpt_0.5.zip)|||
|Multi Band MelGAN |CSMSC|[MB MelGAN-csmsc](https://github.com/PaddlePaddle/PaddleSpeech/tree/develop/examples/csmsc/voc3) | [mb_melgan_baker_ckpt_0.5.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/mb_melgan/mb_melgan_baker_ckpt_0.5.zip) <br>[mb_melgan_baker_finetune_ckpt_0.5.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/mb_melgan/mb_melgan_baker_finetune_ckpt_0.5.zip)|[mb_melgan_baker_static_0.5.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/mb_melgan/mb_melgan_baker_static_0.5.zip) |8.2MB|
|Multi Band MelGAN | CSMSC |[MB MelGAN-csmsc](https://github.com/PaddlePaddle/PaddleSpeech/tree/develop/examples/csmsc/voc3) | [mb_melgan_baker_ckpt_0.5.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/mb_melgan/mb_melgan_baker_ckpt_0.5.zip) <br>[mb_melgan_baker_finetune_ckpt_0.5.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/mb_melgan/mb_melgan_baker_finetune_ckpt_0.5.zip)|[mb_melgan_baker_static_0.5.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/mb_melgan/mb_melgan_baker_static_0.5.zip) |8.2MB|
Style MelGAN | CSMSC |[Style MelGAN-csmsc](https://github.com/PaddlePaddle/PaddleSpeech/tree/develop/examples/csmsc/voc4)|[style_melgan_csmsc_ckpt_0.1.1.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/style_melgan/style_melgan_csmsc_ckpt_0.1.1.zip)| | |
HiFiGAN | CSMSC |[HiFiGAN-csmsc](https://github.com/PaddlePaddle/PaddleSpeech/tree/develop/examples/csmsc/voc5)|[hifigan_csmsc_ckpt_0.1.1.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/hifigan/hifigan_csmsc_ckpt_0.1.1.zip)|[hifigan_csmsc_static_0.1.1.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/hifigan/hifigan_csmsc_static_0.1.1.zip)|50MB|
### Voice Cloning

7
examples/aishell3/tts3/README.md
Unescape View File

@ -72,8 +72,8 @@ Here's the complete help message.
```text
usage: train.py [-h] [--config CONFIG] [--train-metadata TRAIN_METADATA]
[--dev-metadata DEV_METADATA] [--output-dir OUTPUT_DIR]
[--ngpu NGPU] [--verbose VERBOSE] [--phones-dict PHONES_DICT]
[--speaker-dict SPEAKER_DICT]
[--ngpu NGPU] [--phones-dict PHONES_DICT]
[--speaker-dict SPEAKER_DICT] [--voice-cloning VOICE_CLONING]
Train a FastSpeech2 model.
@ -87,11 +87,12 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu=0, use cpu.
--verbose VERBOSE verbose.
--phones-dict PHONES_DICT
phone vocabulary file.
--speaker-dict SPEAKER_DICT
speaker id map file for multiple speaker model.
--voice-cloning VOICE_CLONING
whether training voice cloning model.
```
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.

7
examples/aishell3/voc1/README.md
Unescape View File

@ -67,8 +67,8 @@ Here's the complete help message.
```text
usage: train.py [-h] [--config CONFIG] [--train-metadata TRAIN_METADATA]
[--dev-metadata DEV_METADATA] [--output-dir OUTPUT_DIR]
[--ngpu NGPU] [--verbose VERBOSE] [--batch-size BATCH_SIZE]
[--max-iter MAX_ITER] [--run-benchmark RUN_BENCHMARK]
[--ngpu NGPU] [--batch-size BATCH_SIZE] [--max-iter MAX_ITER]
[--run-benchmark RUN_BENCHMARK]
[--profiler_options PROFILER_OPTIONS]
Train a ParallelWaveGAN model.
@ -83,7 +83,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
benchmark:
arguments related to benchmark.
@ -113,7 +112,6 @@ CUDA_VISIBLE_DEVICES=${gpus} ./local/synthesize.sh ${conf_path} ${train_output_p
usage: synthesize.py [-h] [--generator-type GENERATOR_TYPE] [--config CONFIG]
[--checkpoint CHECKPOINT] [--test-metadata TEST_METADATA]
[--output-dir OUTPUT_DIR] [--ngpu NGPU]
[--verbose VERBOSE]
Synthesize with GANVocoder.
@ -130,7 +128,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
```
1. `--config` parallel wavegan config file. You should use the same config with which the model is trained.

4
examples/csmsc/tts2/README.md
Unescape View File

@ -60,8 +60,7 @@ Here's the complete help message.
```text
usage: train.py [-h] [--config CONFIG] [--train-metadata TRAIN_METADATA]
[--dev-metadata DEV_METADATA] [--output-dir OUTPUT_DIR]
[--ngpu NGPU] [--verbose VERBOSE]
[--use-relative-path USE_RELATIVE_PATH]
[--ngpu NGPU] [--use-relative-path USE_RELATIVE_PATH]
[--phones-dict PHONES_DICT] [--tones-dict TONES_DICT]
Train a Speedyspeech model with a single speaker dataset.
@ -76,7 +75,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
--use-relative-path USE_RELATIVE_PATH
whether use relative path in metadata
--phones-dict PHONES_DICT

6
examples/csmsc/tts2/local/synthesize_e2e.sh
Unescape View File

@ -61,9 +61,9 @@ if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
--am_ckpt=${train_output_path}/checkpoints/${ckpt_name} \
--am_stat=dump/train/feats_stats.npy \
--voc=style_melgan_csmsc \
--voc_config=pwg_baker_ckpt_0.4/pwg_default.yaml \
--voc_ckpt=pwg_baker_ckpt_0.4/pwg_snapshot_iter_400000.pdz \
--voc_stat=pwg_baker_ckpt_0.4/pwg_stats.npy \
--voc_config=style_melgan_csmsc_ckpt_0.1.1/default.yaml \
--voc_ckpt=style_melgan_csmsc_ckpt_0.1.1/snapshot_iter_1500000.pdz \
--voc_stat=style_melgan_csmsc_ckpt_0.1.1/feats_stats.npy \
--lang=zh \
--text=${BIN_DIR}/../sentences.txt \
--output_dir=${train_output_path}/test_e2e \

7
examples/csmsc/tts3/README.md
Unescape View File

@ -63,8 +63,8 @@ Here's the complete help message.
```text
usage: train.py [-h] [--config CONFIG] [--train-metadata TRAIN_METADATA]
[--dev-metadata DEV_METADATA] [--output-dir OUTPUT_DIR]
[--ngpu NGPU] [--verbose VERBOSE] [--phones-dict PHONES_DICT]
[--speaker-dict SPEAKER_DICT]
[--ngpu NGPU] [--phones-dict PHONES_DICT]
[--speaker-dict SPEAKER_DICT] [--voice-cloning VOICE_CLONING]
Train a FastSpeech2 model.
@ -78,11 +78,12 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu=0, use cpu.
--verbose VERBOSE verbose.
--phones-dict PHONES_DICT
phone vocabulary file.
--speaker-dict SPEAKER_DICT
speaker id map file for multiple speaker model.
--voice-cloning VOICE_CLONING
whether training voice cloning model.
```
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.

6
examples/csmsc/tts3/local/synthesize_e2e.sh
Unescape View File

@ -59,9 +59,9 @@ if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
--am_ckpt=${train_output_path}/checkpoints/${ckpt_name} \
--am_stat=dump/train/speech_stats.npy \
--voc=style_melgan_csmsc \
--voc_config=style_melgan_test/default.yaml \
--voc_ckpt=style_melgan_test/snapshot_iter_935000.pdz \
--voc_stat=style_melgan_test/feats_stats.npy \
--voc_config=style_melgan_csmsc_ckpt_0.1.1/default.yaml \
--voc_ckpt=style_melgan_csmsc_ckpt_0.1.1/snapshot_iter_1500000.pdz \
--voc_stat=style_melgan_csmsc_ckpt_0.1.1/feats_stats.npy \
--lang=zh \
--text=${BIN_DIR}/../sentences.txt \
--output_dir=${train_output_path}/test_e2e \

7
examples/csmsc/voc1/README.md
Unescape View File

@ -57,8 +57,8 @@ Here's the complete help message.
```text
usage: train.py [-h] [--config CONFIG] [--train-metadata TRAIN_METADATA]
[--dev-metadata DEV_METADATA] [--output-dir OUTPUT_DIR]
[--ngpu NGPU] [--verbose VERBOSE] [--batch-size BATCH_SIZE]
[--max-iter MAX_ITER] [--run-benchmark RUN_BENCHMARK]
[--ngpu NGPU] [--batch-size BATCH_SIZE] [--max-iter MAX_ITER]
[--run-benchmark RUN_BENCHMARK]
[--profiler_options PROFILER_OPTIONS]
Train a ParallelWaveGAN model.
@ -73,7 +73,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
benchmark:
arguments related to benchmark.
@ -103,7 +102,6 @@ CUDA_VISIBLE_DEVICES=${gpus} ./local/synthesize.sh ${conf_path} ${train_output_p
usage: synthesize.py [-h] [--generator-type GENERATOR_TYPE] [--config CONFIG]
[--checkpoint CHECKPOINT] [--test-metadata TEST_METADATA]
[--output-dir OUTPUT_DIR] [--ngpu NGPU]
[--verbose VERBOSE]
Synthesize with GANVocoder.
@ -120,7 +118,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
```
1. `--config` parallel wavegan config file. You should use the same config with which the model is trained.

5
examples/csmsc/voc3/README.md
Unescape View File

@ -57,7 +57,7 @@ Here's the complete help message.
```text
usage: train.py [-h] [--config CONFIG] [--train-metadata TRAIN_METADATA]
[--dev-metadata DEV_METADATA] [--output-dir OUTPUT_DIR]
[--ngpu NGPU] [--verbose VERBOSE]
[--ngpu NGPU]
Train a Multi-Band MelGAN model.
@ -71,7 +71,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
```
1. `--config` is a config file in yaml format to overwrite the default config, which can be found at `conf/default.yaml`.
@ -88,7 +87,6 @@ CUDA_VISIBLE_DEVICES=${gpus} ./local/synthesize.sh ${conf_path} ${train_output_p
usage: synthesize.py [-h] [--generator-type GENERATOR_TYPE] [--config CONFIG]
[--checkpoint CHECKPOINT] [--test-metadata TEST_METADATA]
[--output-dir OUTPUT_DIR] [--ngpu NGPU]
[--verbose VERBOSE]
Synthesize with GANVocoder.
@ -105,7 +103,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
```
1. `--config` multi band melgan config file. You should use the same config with which the model is trained.

24
examples/csmsc/voc4/README.md
Unescape View File

@ -57,9 +57,9 @@ Here's the complete help message.
```text
usage: train.py [-h] [--config CONFIG] [--train-metadata TRAIN_METADATA]
[--dev-metadata DEV_METADATA] [--output-dir OUTPUT_DIR]
[--ngpu NGPU] [--verbose VERBOSE]
[--ngpu NGPU]
Train a Multi-Band MelGAN model.
Train a Style MelGAN model.
optional arguments:
-h, --help show this help message and exit
@ -71,7 +71,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
```
1. `--config` is a config file in yaml format to overwrite the default config, which can be found at `conf/default.yaml`.
@ -88,7 +87,6 @@ CUDA_VISIBLE_DEVICES=${gpus} ./local/synthesize.sh ${conf_path} ${train_output_p
usage: synthesize.py [-h] [--generator-type GENERATOR_TYPE] [--config CONFIG]
[--checkpoint CHECKPOINT] [--test-metadata TEST_METADATA]
[--output-dir OUTPUT_DIR] [--ngpu NGPU]
[--verbose VERBOSE]
Synthesize with GANVocoder.
@ -105,7 +103,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
```
1. `--config` style melgan config file. You should use the same config with which the model is trained.
@ -113,3 +110,20 @@ optional arguments:
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.
## Pretrained Models
The pretrained model can be downloaded here [style_melgan_csmsc_ckpt_0.1.1.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/style_melgan/style_melgan_csmsc_ckpt_0.1.1.zip).
The static model of Style MelGAN is not available now.
Style MelGAN checkpoint contains files listed below.
```text
hifigan_csmsc_ckpt_0.1.1
├── default.yaml # default config used to train style melgan
├── feats_stats.npy # statistics used to normalize spectrogram when training style melgan
└── snapshot_iter_1500000.pdz # generator parameters of style melgan
```
## Acknowledgement
We adapted some code from https://github.com/kan-bayashi/ParallelWaveGAN.

13
examples/csmsc/voc5/README.md
Unescape View File

@ -57,7 +57,7 @@ Here's the complete help message.
```text
usage: train.py [-h] [--config CONFIG] [--train-metadata TRAIN_METADATA]
[--dev-metadata DEV_METADATA] [--output-dir OUTPUT_DIR]
[--ngpu NGPU] [--verbose VERBOSE]
[--ngpu NGPU]
Train a HiFiGAN model.
@ -71,7 +71,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
```
1. `--config` is a config file in yaml format to overwrite the default config, which can be found at `conf/default.yaml`.
@ -88,7 +87,6 @@ CUDA_VISIBLE_DEVICES=${gpus} ./local/synthesize.sh ${conf_path} ${train_output_p
usage: synthesize.py [-h] [--generator-type GENERATOR_TYPE] [--config CONFIG]
[--checkpoint CHECKPOINT] [--test-metadata TEST_METADATA]
[--output-dir OUTPUT_DIR] [--ngpu NGPU]
[--verbose VERBOSE]
Synthesize with GANVocoder.
@ -105,7 +103,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
```
1. `--config` config file. You should use the same config with which the model is trained.
@ -127,10 +124,10 @@ HiFiGAN checkpoint contains files listed below.
```text
hifigan_csmsc_ckpt_0.1.1
├── default.yaml # default config used to train hifigan
├── feats_stats.npy # generator parameters of hifigan
└── snapshot_iter_2500000.pdz # statistics used to normalize spectrogram when training hifigan
├── default.yaml # default config used to train hifigan
├── feats_stats.npy # statistics used to normalize spectrogram when training hifigan
└── snapshot_iter_2500000.pdz # generator parameters of hifigan
```
## Acknowledgement
We adapted some code from https://github.com/kan-bayashi/ParallelWaveGAN.
We adapted some code from https://github.com/kan-bayashi/ParallelWaveGAN.

44
examples/librispeech/asr1/conf/chunk_conformer.yaml
Unescape View File

@ -47,63 +47,51 @@ data:
dev_manifest: data/manifest.dev
test_manifest: data/manifest.test
collator:
vocab_filepath: data/lang_char/vocab.txt
unit_type: 'spm'
spm_model_prefix: 'data/lang_char/bpe_unigram_5000'
mean_std_filepath: ""
augmentation_config: conf/preprocess.yaml
batch_size: 16
raw_wav: True # use raw_wav or kaldi feature
spectrum_type: fbank #linear, mfcc, fbank
feat_dim: 80
delta_delta: False
dither: 1.0
target_sample_rate: 16000
max_freq: None
n_fft: None
stride_ms: 10.0
window_ms: 25.0
use_dB_normalization: True
target_dB: -20
random_seed: 0
keep_transcription_text: False
sortagrad: True
shuffle_method: batch_shuffle
num_workers: 2
sortagrad: 0 # Feed samples from shortest to longest ; -1: enabled for all epochs, 0: disabled, other: enabled for 'other' epochs
batch_size: 16
maxlen_in: 512 # if input length > maxlen-in, batchsize is automatically reduced
maxlen_out: 150 # if output length > maxlen-out, batchsize is automatically reduced
minibatches: 0 # for debug
batch_count: auto
batch_bins: 0
batch_frames_in: 0
batch_frames_out: 0
batch_frames_inout: 0
augmentation_config: conf/preprocess.yaml
num_workers: 0
subsampling_factor: 1
num_encs: 1
training:
n_epoch: 240
n_epoch: 120
accum_grad: 8
global_grad_clip: 5.0
optim: adam
optim_conf:
lr: 0.001
weight_decay: 1e-06
weight_decay: 1e-06
scheduler: warmuplr
scheduler_conf:
warmup_steps: 25000
lr_decay: 1.0
log_interval: 100
checkpoint:
kbest_n: 50
latest_n: 5
decoding:
batch_size: 128
error_rate_type: wer
decoding_method: attention # 'attention', 'ctc_greedy_search', 'ctc_prefix_beam_search', 'attention_rescoring'
lang_model_path: data/lm/common_crawl_00.prune01111.trie.klm
alpha: 2.5
beta: 0.3
beam_size: 10
cutoff_prob: 1.0
cutoff_top_n: 0
num_proc_bsearch: 8
ctc_weight: 0.5 # ctc weight for attention rescoring decode mode.
decoding_chunk_size: -1 # decoding chunk size. Defaults to -1.
# <0: for decoding, use full chunk.

35
examples/librispeech/asr1/conf/chunk_transformer.yaml
Unescape View File

@ -34,36 +34,35 @@ model:
lsm_weight: 0.1 # label smoothing option
length_normalized_loss: false
data:
train_manifest: data/manifest.train
dev_manifest: data/manifest.dev
test_manifest: data/manifest.test
collator:
vocab_filepath: data/lang_char/vocab.txt
unit_type: 'spm'
spm_model_prefix: 'data/lang_char/bpe_unigram_5000'
mean_std_filepath: ""
augmentation_config: conf/preprocess.yaml
batch_size: 64
raw_wav: True # use raw_wav or kaldi feature
spectrum_type: fbank #linear, mfcc, fbank
feat_dim: 80
delta_delta: False
dither: 1.0
target_sample_rate: 16000
max_freq: None
n_fft: None
stride_ms: 10.0
window_ms: 25.0
use_dB_normalization: True
target_dB: -20
random_seed: 0
keep_transcription_text: False
sortagrad: True
shuffle_method: batch_shuffle
num_workers: 2
sortagrad: 0 # Feed samples from shortest to longest ; -1: enabled for all epochs, 0: disabled, other: enabled for 'other' epochs
batch_size: 64
maxlen_in: 512 # if input length > maxlen-in, batchsize is automatically reduced
maxlen_out: 150 # if output length > maxlen-out, batchsize is automatically reduced
minibatches: 0 # for debug
batch_count: auto
batch_bins: 0
batch_frames_in: 0
batch_frames_out: 0
batch_frames_inout: 0
augmentation_config: conf/preprocess.yaml
num_workers: 0
subsampling_factor: 1
num_encs: 1
training:
@ -101,6 +100,4 @@ decoding:
# >0: for decoding, use fixed chunk size as set.
# 0: used for training, it's prohibited here.
num_decoding_left_chunks: -1 # number of left chunks for decoding. Defaults to -1.
simulate_streaming: true # simulate streaming inference. Defaults to False.
simulate_streaming: true # simulate streaming inference. Defaults to False.

32
examples/librispeech/asr1/conf/conformer.yaml
Unescape View File

@ -34,6 +34,7 @@ model:
# hybrid CTC/attention
model_conf:
ctc_weight: 0.3
ctc_grad_norm_type: null
lsm_weight: 0.1 # label smoothing option
length_normalized_loss: false
@ -50,25 +51,24 @@ collator:
spm_model_prefix: 'data/lang_char/bpe_unigram_5000'
mean_std_filepath: ""
augmentation_config: conf/preprocess.yaml
batch_size: 16
raw_wav: True # use raw_wav or kaldi feature
spectrum_type: fbank #linear, mfcc, fbank
feat_dim: 80
delta_delta: False
dither: 1.0
target_sample_rate: 16000
max_freq: None
n_fft: None
stride_ms: 10.0
window_ms: 25.0
use_dB_normalization: True
target_dB: -20
random_seed: 0
keep_transcription_text: False
sortagrad: True
shuffle_method: batch_shuffle
num_workers: 2
sortagrad: 0 # Feed samples from shortest to longest ; -1: enabled for all epochs, 0: disabled, other: enabled for 'other' epochs
batch_size: 16
maxlen_in: 512 # if input length > maxlen-in, batchsize is automatically reduced
maxlen_out: 150 # if output length > maxlen-out, batchsize is automatically reduced
minibatches: 0 # for debug
batch_count: auto
batch_bins: 0
batch_frames_in: 0
batch_frames_out: 0
batch_frames_inout: 0
augmentation_config: conf/preprocess.yaml
num_workers: 0
subsampling_factor: 1
num_encs: 1
training:
n_epoch: 70

29
examples/librispeech/asr1/conf/transformer.yaml
Unescape View File

@ -51,24 +51,23 @@ collator:
spm_model_prefix: 'data/lang_char/bpe_unigram_5000'
mean_std_filepath: ""
augmentation_config: conf/preprocess.yaml
batch_size: 32
raw_wav: True # use raw_wav or kaldi feature
spectrum_type: fbank #linear, mfcc, fbank
feat_dim: 80
delta_delta: False
dither: 1.0
target_sample_rate: 16000
max_freq: None
n_fft: None
stride_ms: 10.0
window_ms: 25.0
use_dB_normalization: True
target_dB: -20
random_seed: 0
keep_transcription_text: False
sortagrad: True
shuffle_method: batch_shuffle
num_workers: 2
sortagrad: 0 # Feed samples from shortest to longest ; -1: enabled for all epochs, 0: disabled, other: enabled for 'other' epochs
batch_size: 32
maxlen_in: 512 # if input length > maxlen-in, batchsize is automatically reduced
maxlen_out: 150 # if output length > maxlen-out, batchsize is automatically reduced
minibatches: 0 # for debug
batch_count: auto
batch_bins: 0
batch_frames_in: 0
batch_frames_out: 0
batch_frames_inout: 0
augmentation_config: conf/preprocess.yaml
num_workers: 0
subsampling_factor: 1
num_encs: 1
training:

8
examples/ljspeech/tts1/README.md
Unescape View File

@ -55,7 +55,7 @@ Here's the complete help message.
```text
usage: train.py [-h] [--config CONFIG] [--train-metadata TRAIN_METADATA]
[--dev-metadata DEV_METADATA] [--output-dir OUTPUT_DIR]
[--ngpu NGPU] [--verbose VERBOSE] [--phones-dict PHONES_DICT]
[--ngpu NGPU] [--phones-dict PHONES_DICT]
Train a TransformerTTS model with LJSpeech TTS dataset.
@ -69,7 +69,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
--phones-dict PHONES_DICT
phone vocabulary file.
```
@ -103,7 +102,7 @@ usage: synthesize.py [-h] [--transformer-tts-config TRANSFORMER_TTS_CONFIG]
[--waveflow-checkpoint WAVEFLOW_CHECKPOINT]
[--phones-dict PHONES_DICT]
[--test-metadata TEST_METADATA] [--output-dir OUTPUT_DIR]
[--ngpu NGPU] [--verbose VERBOSE]
[--ngpu NGPU]
Synthesize with transformer tts & waveflow.
@ -127,7 +126,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
```
`./local/synthesize_e2e.sh` calls `${BIN_DIR}/synthesize_e2e.py`, which can synthesize waveform from text file.
```bash
@ -142,7 +140,6 @@ usage: synthesize_e2e.py [-h]
[--waveflow-checkpoint WAVEFLOW_CHECKPOINT]
[--phones-dict PHONES_DICT] [--text TEXT]
[--output-dir OUTPUT_DIR] [--ngpu NGPU]
[--verbose VERBOSE]
Synthesize with transformer tts & waveflow.
@ -165,7 +162,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
```
1. `--transformer-tts-config`, `--transformer-tts-checkpoint`, `--transformer-tts-stat` and `--phones-dict` are arguments for transformer_tts, which correspond to the 4 files in the transformer_tts pretrained model.
2. `--waveflow-config`, `--waveflow-checkpoint` are arguments for waveflow, which correspond to the 2 files in the waveflow pretrained model.

7
examples/ljspeech/tts3/README.md
Unescape View File

@ -62,8 +62,8 @@ Here's the complete help message.
```text
usage: train.py [-h] [--config CONFIG] [--train-metadata TRAIN_METADATA]
[--dev-metadata DEV_METADATA] [--output-dir OUTPUT_DIR]
[--ngpu NGPU] [--verbose VERBOSE] [--phones-dict PHONES_DICT]
[--speaker-dict SPEAKER_DICT]
[--ngpu NGPU] [--phones-dict PHONES_DICT]
[--speaker-dict SPEAKER_DICT] [--voice-cloning VOICE_CLONING]
Train a FastSpeech2 model.
@ -77,11 +77,12 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu=0, use cpu.
--verbose VERBOSE verbose.
--phones-dict PHONES_DICT
phone vocabulary file.
--speaker-dict SPEAKER_DICT
speaker id map file for multiple speaker model.
--voice-cloning VOICE_CLONING
whether training voice cloning model.
```
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.

7
examples/ljspeech/voc1/README.md
Unescape View File

@ -57,8 +57,8 @@ Here's the complete help message.
```text
usage: train.py [-h] [--config CONFIG] [--train-metadata TRAIN_METADATA]
[--dev-metadata DEV_METADATA] [--output-dir OUTPUT_DIR]
[--ngpu NGPU] [--verbose VERBOSE] [--batch-size BATCH_SIZE]
[--max-iter MAX_ITER] [--run-benchmark RUN_BENCHMARK]
[--ngpu NGPU] [--batch-size BATCH_SIZE] [--max-iter MAX_ITER]
[--run-benchmark RUN_BENCHMARK]
[--profiler_options PROFILER_OPTIONS]
Train a ParallelWaveGAN model.
@ -73,7 +73,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
benchmark:
arguments related to benchmark.
@ -103,7 +102,6 @@ CUDA_VISIBLE_DEVICES=${gpus} ./local/synthesize.sh ${conf_path} ${train_output_p
usage: synthesize.py [-h] [--generator-type GENERATOR_TYPE] [--config CONFIG]
[--checkpoint CHECKPOINT] [--test-metadata TEST_METADATA]
[--output-dir OUTPUT_DIR] [--ngpu NGPU]
[--verbose VERBOSE]
Synthesize with GANVocoder.
@ -120,7 +118,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
```
1. `--config` parallel wavegan config file. You should use the same config with which the model is trained.

7
examples/vctk/tts3/README.md
Unescape View File

@ -65,8 +65,8 @@ Here's the complete help message.
```text
usage: train.py [-h] [--config CONFIG] [--train-metadata TRAIN_METADATA]
[--dev-metadata DEV_METADATA] [--output-dir OUTPUT_DIR]
[--ngpu NGPU] [--verbose VERBOSE] [--phones-dict PHONES_DICT]
[--speaker-dict SPEAKER_DICT]
[--ngpu NGPU] [--phones-dict PHONES_DICT]
[--speaker-dict SPEAKER_DICT] [--voice-cloning VOICE_CLONING]
Train a FastSpeech2 model.
@ -80,11 +80,12 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu=0, use cpu.
--verbose VERBOSE verbose.
--phones-dict PHONES_DICT
phone vocabulary file.
--speaker-dict SPEAKER_DICT
speaker id map file for multiple speaker model.
--voice-cloning VOICE_CLONING
whether training voice cloning model.
```
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.

7
examples/vctk/voc1/README.md
Unescape View File

@ -62,8 +62,8 @@ Here's the complete help message.
```text
usage: train.py [-h] [--config CONFIG] [--train-metadata TRAIN_METADATA]
[--dev-metadata DEV_METADATA] [--output-dir OUTPUT_DIR]
[--ngpu NGPU] [--verbose VERBOSE] [--batch-size BATCH_SIZE]
[--max-iter MAX_ITER] [--run-benchmark RUN_BENCHMARK]
[--ngpu NGPU] [--batch-size BATCH_SIZE] [--max-iter MAX_ITER]
[--run-benchmark RUN_BENCHMARK]
[--profiler_options PROFILER_OPTIONS]
Train a ParallelWaveGAN model.
@ -78,7 +78,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
benchmark:
arguments related to benchmark.
@ -108,7 +107,6 @@ CUDA_VISIBLE_DEVICES=${gpus} ./local/synthesize.sh ${conf_path} ${train_output_p
usage: synthesize.py [-h] [--generator-type GENERATOR_TYPE] [--config CONFIG]
[--checkpoint CHECKPOINT] [--test-metadata TEST_METADATA]
[--output-dir OUTPUT_DIR] [--ngpu NGPU]
[--verbose VERBOSE]
Synthesize with GANVocoder.
@ -125,7 +123,6 @@ optional arguments:
--output-dir OUTPUT_DIR
output dir.
--ngpu NGPU if ngpu == 0, use cpu.
--verbose VERBOSE verbose.
```
1. `--config` parallel wavegan config file. You should use the same config with which the model is trained.

3
paddlespeech/t2s/exps/gan_vocoder/style_melgan/train.py
Unescape View File

@ -223,8 +223,7 @@ def train_sp(args, config):
def main():
# parse args and config and redirect to train_sp
parser = argparse.ArgumentParser(
description="Train a Multi-Band MelGAN model.")
parser = argparse.ArgumentParser(description="Train a Style MelGAN model.")
parser.add_argument(
"--config", type=str, help="config file to overwrite default config.")
parser.add_argument("--train-metadata", type=str, help="training data.")

1
paddlespeech/t2s/exps/speedyspeech/train.py
Unescape View File

@ -168,7 +168,6 @@ def main():
parser.add_argument("--output-dir", type=str, help="output dir.")
parser.add_argument(
"--ngpu", type=int, default=1, help="if ngpu == 0, use cpu.")
parser.add_argument("--verbose", type=int, default=1, help="verbose.")
def str2bool(str):
return True if str.lower() == 'true' else False

2
paddlespeech/t2s/frontend/zh_frontend.py
Unescape View File

@ -106,7 +106,7 @@ class Frontend():
for seg in segments:
phones = []
# Replace all English words in the sentence
seg = re.sub('[a-zA-Z]+','',seg)
seg = re.sub('[a-zA-Z]+', '', seg)
seg_cut = psg.lcut(seg)
initials = []
finals = []

12
paddlespeech/t2s/models/fastspeech2/fastspeech2.py
Unescape View File

@ -942,7 +942,12 @@ class StyleFastSpeech2Inference(FastSpeech2Inference):
"""
spk_id = paddle.to_tensor(spk_id)
normalized_mel, d_outs, p_outs, e_outs = self.acoustic_model.inference(
text, durations=None, pitch=None, energy=None, spk_emb=spk_emb, spk_id=spk_id)
text,
durations=None,
pitch=None,
energy=None,
spk_emb=spk_emb,
spk_id=spk_id)
# priority: groundtruth > scale/bias > previous output
# set durations
if isinstance(durations, np.ndarray):
@ -995,9 +1000,8 @@ class StyleFastSpeech2Inference(FastSpeech2Inference):
pitch=pitch,
energy=energy,
use_teacher_forcing=True,
spk_emb=spk_emb,
spk_id=spk_id
)
spk_emb=spk_emb,
spk_id=spk_id)
logmel = self.normalizer.inverse(normalized_mel)
return logmel

3
paddlespeech/t2s/models/melgan/style_melgan.py
Unescape View File

@ -188,7 +188,8 @@ class StyleMelGANGenerator(nn.Layer):
try:
if layer:
nn.utils.remove_weight_norm(layer)
except ValueError:
# add AttributeError to bypass https://github.com/PaddlePaddle/Paddle/issues/38532 temporarily
except (ValueError, AttributeError):
pass
self.apply(_remove_weight_norm)

6
paddlespeech/t2s/modules/tade_res_block.py
Unescape View File

@ -33,7 +33,11 @@ class TADELayer(nn.Layer):
"""Initilize TADE layer."""
super().__init__()
self.norm = nn.InstanceNorm1D(
in_channels, momentum=0.1, data_format="NCL")
in_channels,
momentum=0.1,
data_format="NCL",
weight_attr=False,
bias_attr=False)
self.aux_conv = nn.Sequential(
nn.Conv1D(
aux_channels,

Write Preview
Loading…
Cancel
Save