模型介绍
内容详情
替代品
模型简介
该模型能够将韩语文本转换为自然语音，适用于语音合成应用。
模型特点
高质量语音合成
采用JETS架构，能够生成自然流畅的韩语语音
端到端训练
整个系统采用端到端方式训练，简化了传统TTS系统的复杂流程
韩语优化
专门针对韩语语音特性进行优化，使用g2pk进行韩语文本处理
模型能力
韩语文本转语音
语音合成
使用案例
语音助手
韩语语音助手
为韩语语音助手提供自然语音输出
有声读物
韩语有声内容生成
将韩语文本内容自动转换为语音
标签:
- espnet
- 音频
- 文本转语音 语言: 韩语 数据集:
- kss 许可证: cc-by-4.0
ESPnet2 TTS 模型
imdanboy/kss_tts_train_jets_raw_phn_null_g2pk_train.total_count.ave
此模型由 satoshi.2020 使用 espnet 中的 kss 配方训练而成。
演示：如何在 ESPnet2 中使用
cd espnet
git checkout 047d0c474c18a87c205e566948410be16787e477
pip install -e .
cd egs2/kss/tts1
./run.sh --skip_data_prep false --skip_train true --download_model imdanboy/kss_tts_train_jets_raw_phn_null_g2pk_train.total_count.ave
TTS 配置
展开
配置: conf/tuning/train_jets.yaml
print_config: false
日志级别: INFO
dry_run: false
iterator_type: sequence
输出目录: exp/tts_train_jets_raw_phn_null_g2pk
ngpu: 1
种子: 777
num_workers: 4
num_att_plot: 3
dist_backend: nccl
dist_init_method: env://
dist_world_size: 4
dist_rank: 0
local_rank: 0
dist_master_addr: localhost
dist_master_port: 52809
dist_launcher: null
multiprocessing_distributed: true
unused_parameters: true
sharded_ddp: false
cudnn_enabled: true
cudnn_benchmark: false
cudnn_deterministic: false
collect_stats: false
write_collected_feats: false
max_epoch: 1000
patience: null
val_scheduler_criterion:
- valid
- loss
early_stopping_criterion:
- valid
- loss
- min
best_model_criterion:
- - valid
- text2mel_loss
- min
- - train
- text2mel_loss
- min
- - train
- total_count
- max
keep_nbest_models: 5
nbest_averaging_interval: 0
grad_clip: -1
grad_clip_type: 2.0
grad_noise: false
accum_grad: 1
no_forward_run: false
resume: true
train_dtype: float32
use_amp: false
log_interval: 50
use_matplotlib: true
use_tensorboard: true
use_wandb: false
wandb_project: null
wandb_id: null
wandb_entity: null
wandb_name: null
wandb_model_log_interval: -1
detect_anomaly: false
pretrain_path: null
init_param: []
ignore_init_mismatch: false
freeze_param: []
num_iters_per_epoch: 1000
batch_size: 20
valid_batch_size: null
batch_bins: 2000000
valid_batch_bins: null
train_shape_file:
- exp/tts_stats_raw_phn_null_g2pk/train/text_shape.phn
- exp/tts_stats_raw_phn_null_g2pk/train/speech_shape
valid_shape_file:
- exp/tts_stats_raw_phn_null_g2pk/valid/text_shape.phn
- exp/tts_stats_raw_phn_null_g2pk/valid/speech_shape
batch_type: numel
valid_batch_type: null
fold_length:
- 150
- 204800
sort_in_batch: descending
sort_batch: descending
multiple_iterator: false
chunk_length: 500
chunk_shift_ratio: 0.5
num_cache_chunks: 1024
train_data_path_and_name_and_type:
- - dump/raw/tr_no_dev/text
- text
- text
- - dump/raw/tr_no_dev/wav.scp
- speech
- sound
- - exp/tts_stats_raw_phn_null_g2pk/train/collect_feats/pitch.scp
- pitch
- npy
- - exp/tts_stats_raw_phn_null_g2pk/train/collect_feats/energy.scp
- energy
- npy
valid_data_path_and_name_and_type:
- - dump/raw/dev/text
- text
- text
- - dump/raw/dev/wav.scp
- speech
- sound
- - exp/tts_stats_raw_phn_null_g2pk/valid/collect_feats/pitch.scp
- pitch
- npy
- - exp/tts_stats_raw_phn_null_g2pk/valid/collect_feats/energy.scp
- energy
- npy
allow_variable_data_keys: false
max_cache_size: 0.0
max_cache_fd: 32
valid_max_cache_size: null
optim: adamw
optim_conf:
lr: 0.0002
betas:
- 0.8
- 0.99
eps: 1.0e-09
weight_decay: 0.0
scheduler: exponentiallr
scheduler_conf:
gamma: 0.999875
optim2: adamw
optim2_conf:
lr: 0.0002
betas:
- 0.8
- 0.99
eps: 1.0e-09
weight_decay: 0.0
scheduler2: exponentiallr
scheduler2_conf:
gamma: 0.999875
generator_first: true
token_list:
- <blank>
- <unk>
- ''
- ᅡ
- ᅵ
- ᄋ
- ᅳ
- ᄀ
- ᅥ
- ᄂ
- ᆫ
- ᄅ
- ᄌ
- ᄉ
- ᅩ
- ᆯ
- ᄆ
- .
- ᅮ
- ᄃ
- ᄒ
- ᅦ
- ᆼ
- ᅢ
- ᄇ
- ᅭ
- ᅧ
- ᄊ
- ᆷ
- ᄄ
- ᆮ
- ᄎ
- ᄁ
- ᆨ
- ᄑ
- ᄐ
- ᅪ
- ᄏ
- '?'
- ᄍ
- ᆸ
- ᅬ
- ᅣ
- ᅴ
- ᅯ
- ᅨ
- ᄈ
- ᅱ
- ᅲ
- ᅫ
- ','
- '!'
- ᅤ
- ':'
- ᅰ
- ''''
- '-'
- '"'
- /
- I
- M
- F
- E
- S
- C
- A
- B
- ㅇ
- <sos/eos>
odim: null
model_conf: {}
use_preprocessor: true
token_type: phn
bpemodel: null
non_linguistic_symbols: null
cleaner: null
g2p: g2pk
feats_extract: fbank
feats_extract_conf:
n_fft: 1024
hop_length: 256
win_length: null
fs: 24000
fmin: 0
fmax: null
n_mels: 80
normalize: global_mvn
normalize_conf:
stats_file: exp/tts_stats_raw_phn_null_g2pk/train/feats_stats.npz
tts: jets
tts_conf:
generator_type: jets_generator
generator_params:
adim: 256
aheads: 2
elayers: 4
eunits: 1024
dlayers: 4
dunits: 1024
positionwise_layer_type: conv1d
positionwise_conv_kernel_size: 3
duration_predictor_layers: 2
duration_predictor_chans: 256
duration_predictor_kernel_size: 3
use_masking: true
encoder_normalize_before: true
decoder_normalize_before: true
encoder_type: transformer
decoder_type: transformer
conformer_rel_pos_type: latest
conformer_pos_enc_layer_type: rel_pos
conformer_self_attn_layer_type: rel_selfattn
conformer_activation_type: swish
use_macaron_style_in_conformer: true
use_cnn_in_conformer: true
conformer_enc_kernel_size: 7
conformer_dec_kernel_size: 31
init_type: xavier_uniform
transformer_enc_dropout_rate: 0.2
transformer_enc_positional_dropout_rate: 0.2
transformer_enc_attn_dropout_rate: 0.2
transformer_dec_dropout_rate: 0.2
transformer_dec_positional_dropout_rate: 0.2
transformer_dec_attn_dropout_rate: 0.2
pitch_predictor_layers: 5
pitch_predictor_chans: 256
pitch_predictor_kernel_size: 5
pitch_predictor_dropout: 0.5
pitch_embed_kernel_size: 1
pitch_embed_dropout: 0.0
stop_gradient_from_pitch_predictor: true
energy_predictor_layers: 2
energy_predictor_chans: 256
energy_predictor_kernel_size: 3
energy_predictor_dropout: 0.5
energy_embed_kernel_size: 1
energy_embed_dropout: 0.0
stop_gradient_from_energy_predictor: false
generator_out_channels: 1
generator_channels: 512
generator_global_channels: -1
generator_kernel_size: 7
generator_upsample_scales:
- 8
- 8
- 2
- 2
generator_upsample_kernel_sizes:
- 16
- 16
- 4
- 4
generator_resblock_kernel_sizes:
- 3
- 7
- 11
generator_resblock_dilations:
- - 1
- 3
- 5
- - 1
- 3
- 5
- - 1
- 3
- 5
generator_use_additional_convs: true
generator_bias: true
generator_nonlinear_activation: LeakyReLU
generator_nonlinear_activation_params:
negative_slope: 0.1
generator_use_weight_norm: true
segment_size: 64
idim: 69
odim: 80
discriminator_type: hifigan_multi_scale_multi_period_discriminator
discriminator_params:
scales: 1
scale_downsample_pooling: AvgPool1d
scale_downsample_pooling_params:
kernel_size: 4
stride: 2
padding: 2
scale_discriminator_params:
in_channels: 1
out_channels: 1
kernel_sizes:
- 15
- 41
- 5
- 3
channels: 128
max_downsample_channels: 1024
max_groups: 16
bias: true
downsample_scales:
- 2
- 2
- 4
- 4
- 1
nonlinear_activation: LeakyReLU
nonlinear_activation_params:
negative_slope: 0.1
use_weight_norm: true
use_spectral_norm: false
follow_official_norm: false
periods:
- 2
- 3
- 5
- 7
- 11
period_discriminator_params:
in_channels: 1
out_channels: 1
kernel_sizes:
- 5
- 3
channels: 32
downsample_scales:
- 3
- 3
- 3
- 3
- 1
max_downsample_channels: 1024
bias: true
nonlinear_activation: LeakyReLU
nonlinear_activation_params:
negative_slope: 0.1
use_weight_norm: true
use_spectral_norm: false
generator_adv_loss_params:
average_by_discriminators: false
loss_type: mse
discriminator_adv_loss_params:
average_by_discriminators: false
loss_type: mse
feat_match_loss_params:
average_by_discriminators: false
average_by_layers: false
include_final_outputs: true
mel_loss_params:
fs: 24000
n_fft: 1024
hop_length: 256
win_length: null
window: hann
n_mels: 80
fmin: 0
fmax: null
log_base: null
lambda_adv: 1.0
lambda_mel: 45.0
lambda_feat_match: 2.0
lambda_var: 1.0
lambda_align: 2.0
sampling_rate: 24000
cache_generator_outputs: true
pitch_extract: dio
pitch_extract_conf:
reduction_factor: 1
use_token_averaged_f0: false
fs: 24000
n_fft: 1024
hop_length: 256
f0max: 400
f0min: 80
pitch_normalize: global_mvn
pitch_normalize_conf:
stats_file: exp/tts_stats_raw_phn_null_g2pk/train/pitch_stats.npz
energy_extract: energy
energy_extract_conf:
reduction_factor: 1
use_token_averaged_energy: false
fs: 24000
n_fft: 1024
hop_length: 256
win_length: null
energy_normalize: global_mvn
energy_normalize_conf:
stats_file: exp/tts_stats_raw_phn_null_g2pk/train/energy_stats.npz
required:
- output_dir
- token_list
version: '202204'
distributed: true
引用 ESPnet
@inproceedings{watanabe2018espnet,
作者={Shinji Watanabe 和 Takaaki Hori 和 Shigeki Karita 和 Tomoki Hayashi 和 Jiro Nishitoba 和 Yuya Unno 和 Nelson Yalta 和 Jahn Heymann 和 Matthew Wiesner 和 Nanxin Chen 和 Adithya Renduchintala 和 Tsubasa Ochiai},
标题={{ESPnet}: 端到端语音处理工具包},
年份={2018},
会议={Interspeech 会议论文集},
页码={2207--2211},
doi={10.21437/Interspeech.2018-1456},
url={http://dx.doi.org/10.21437/Interspeech.2018-1456}
}
@inproceedings{hayashi2020espnet,
标题={{Espnet-TTS}: 统一、可复现且可集成的开源端到端文本转语音工具包},
作者={Hayashi, Tomoki 和 Yamamoto, Ryuichi 和 Inoue, Katsuki 和 Yoshimura, Takenori 和 Watanabe, Shinji 和 Toda, Tomoki 和 Takeda, Kazuya 和 Zhang, Yu 和 Tan, Xu},
会议={IEEE 国际声学、语音与信号处理会议 (ICASSP) 论文集},
页码={7654--7658},
年份={2020},
组织={IEEE}
}
或 arXiv:
@misc{watanabe2018espnet,
标题={ESPnet: 端到端语音处理工具包},
作者={Shinji Watanabe 和 Takaaki Hori 和 Shigeki Karita 和 Tomoki Hayashi 和 Jiro Nishitoba 和 Yuya Unno 和 Nelson Yalta 和 Jahn Heymann 和 Matthew Wiesner 和 Nanxin Chen 和 Adithya Renduchintala 和 Tsubasa Ochiai},
年份={2018},
eprint={1804.00015},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
Kokoro 82M
Apache-2.0
Kokoro是一款拥有8200万参数的开源文本转语音（TTS）模型，以其轻量级架构和高音质著称，同时具备快速和成本效益高的特点。
语音合成
英语
K
hexgrad
2.0M
4,155
XTTS V2
其他
ⓍTTS是一款革命性的语音生成模型，仅需6秒音频片段即可实现跨语言音色克隆，支持17种语言。
语音合成
X
coqui
1.7M
2,630
F5 TTS
F5-TTS 是一个基于流匹配的语音合成模型，专注于流畅且忠实的语音合成，特别适用于童话讲述等场景。
语音合成
F
SWivid
851.49k
1,000
Bigvgan V2 22khz 80band 256x
MIT
BigVGAN是基于大规模训练的通用神经声码器，能够从梅尔频谱生成高质量音频波形。
语音合成
B
nvidia
503.23k
16
Speecht5 Tts
MIT
基于LibriTTS数据集微调的SpeechT5语音合成(文本转语音)模型，支持高质量的文本转语音转换。
语音合成
Transformers
TransformersS
microsoft
113.83k
760
Dia 1.6B
Apache-2.0
Dia是由Nari实验室开发的16亿参数文本转语音模型，能够直接从文本生成高度逼真的对话，支持情感和语调控制，并能生成非语言交流内容。
语音合成
Safetensors
Safetensors 英语
D
nari-labs
80.28k
1,380
Csm 1b
Apache-2.0
CSM是Sesame开发的10亿参数规模语音生成模型，可根据文本和音频输入生成RVQ音频编码
语音合成 Safetensors
Safetensors 英语
C
sesame
65.03k
1,950
Kokoro 82M V1.1 Zh
Apache-2.0
Kokoro 是一个开放权重的小型但功能强大的文本转语音（TTS）模型系列，新增了来自专业数据集的100名中文说话人数据。
语音合成
K
hexgrad
51.56k
112
Indic Parler Tts
Apache-2.0
Indic Parler-TTS 是 Parler-TTS Mini 的多语言印度语言扩展版本，支持21种语言，包括多种印度语言和英语。
语音合成 Transformers
Transformers 支持多种语言
I
ai4bharat
43.59k
124
Bark
MIT
Bark是由Suno创建的基于Transformer的文本转音频模型，能生成高度逼真的多语言语音、音乐、背景噪音和简单音效。
语音合成 Transformers
Transformers 支持多种语言
B
suno
35.72k
1,326
精选推荐AI模型
Llama 3 Typhoon V1.5x 8b Instruct
专为泰语设计的80亿参数指令模型，性能媲美GPT-3.5-turbo，优化了应用场景、检索增强生成、受限生成和推理任务
大型语言模型 Transformers
Transformers 支持多种语言
L
scb10x
3,269
16
Cadet Tiny
Openrail
Cadet-Tiny是一个基于SODA数据集训练的超小型对话模型，专为边缘设备推理设计，体积仅为Cosmo-3B模型的2%左右。
对话系统 Transformers
Transformers 英语
C
ToddGoldfarb
2,691
6
Roberta Base Chinese Extractive Qa
基于RoBERTa架构的中文抽取式问答模型，适用于从给定文本中提取答案的任务。
问答系统
中文
R
uer
2,694
98
%20--%3e%3cdefs%3e%3cstyle%3e%20.st0%20{%20fill:%20none;%20}%20.st1%20{%20fill:%20%235ce6cf;%20}%20.st2%20{%20fill:%20%230080ff;%20}%20.st3%20{%20clip-path:%20url(%23clippath);%20}%20%3c/style%3e%3cclipPath%20id='clippath'%3e%3crect%20class='st0'%20y='0'%20width='120'%20height='32'/%3e%3c/clipPath%3e%3c/defs%3e%3cg%20class='st3'%3e%3cg%3e%3cpath%20class='st1'%20d='M22.4,32H3.2c-1,0-2-.5-2.6-1.4-.6-.9-.7-2-.4-2.9L9.8,2.1C10.3.8,11.5,0,12.8,0s2.5.8,3,2.1l9.6,25.6c.4,1,.2,2.1-.4,2.9-.6.9-1.6,1.4-2.6,1.4ZM7.8,25.6h10l-5-13.3-5,13.3Z'/%3e%3cpath%20class='st2'%20d='M33.6,32c-1.3,0-2.5-.8-3-2.1L21,4.3c-.6-1.7.2-3.5,1.9-4.1,1.7-.6,3.5.2,4.1,1.9l9.6,25.6c.6,1.7-.2,3.5-1.9,4.1-.4.1-.8.2-1.1.2h0Z'/%3e%3cpath%20d='M71.8,9.1h-6.9v4.6h6.9c2.5,0,4.6,2,4.6,4.6h-6.9c-3.8,0-6.9,3.1-6.9,6.9s3.1,6.9,6.9,6.9h2.3c1.7,0,3.2-.4,4.6-1.2v1.2h4.6v-13.7c0-5-4.1-9.1-9.1-9.1h0ZM71.8,27.4h-2.3c-1.3,0-2.3-1-2.3-2.3s1-2.3,2.3-2.3h6.9c0,2.5-2,4.6-4.6,4.6h0Z'/%3e%3cpath%20d='M52.3,9.1h-4.6V0h-4.6V32h9.1c5,0,9.1-4.1,9.1-9.1v-4.6c0-5.1-4.1-9.1-9.1-9.1h0ZM56.9,22.9c0,2.5-2,4.6-4.6,4.6h-4.6v-13.7h4.6c2.5,0,4.6,2,4.6,4.6v4.6h0Z'/%3e%3cpath%20d='M110.9,13.7h9.1v-4.6h-9.1c-5.1,0-9.1,4.1-9.1,9.1v4.6c0,5.1,4.1,9.1,9.1,9.1h9.1v-4.6h-9.1c-2.5,0-4.6-2-4.6-4.6h13.7v-4.6h-13.7c0-2.5,2-4.6,4.6-4.6h0Z'/%3e%3cpath%20d='M93.6,18.3h-4.6c-1.3,0-2.3-1-2.3-2.3s1-2.3,2.3-2.3h10.3v-4.6h-10.3c-3.8,0-6.9,3.1-6.9,6.9s3.1,6.9,6.9,6.9h4.6c1.3,0,2.3,1,2.3,2.3s-1,2.3-2.3,2.3h-10.3v4.6h10.3c3.8,0,6.9-3.1,6.9-6.9s-3.1-6.9-6.9-6.9Z'/%3e%3c/g%3e%3c/g%3e%3c/svg%3e)
AIbase是一个专注于MCP服务的平台，为AI开发者提供高质量的模型上下文协议服务，助力AI应用开发。
简体中文