KAN-TTS

KAN-TTS copied to clipboard

SAM-BERT流式

1

请问下SAM-BERT流式合成的方案要怎么实现？

FB-wh

Need python 3.9 & python 3.10 supports

2

```shell pip install git+https://github.com/alibaba-damo-academy/KAN-TTS INFO: pip is looking at multiple versions of kantts to determine which version is compatible with other requirements. This could take a while. ERROR: Package 'kantts'...

nne998

训练步数越多 反而效果不好 有杂音

1

训练步数越多 反而效果不好 有杂音 有朋友遇到么

dushaofan

纯英文效果不好，怎么破？一个单词一个单词往外蹦。。

3

需要合成纯英文，必须用英文音频训练吗？现在我用中文音频训练后，然后推理英文的时候，效果不好，一个单词一个单词往外蹦。。 然后我想用纯英文音频训练，结果第一步数据预处理就出错，无法自动标注 大佬，怎么破？

stevin-dong

https://modelscope.oss-cn-beijing.aliyuncs.com/releases/repo.html 上又只有0.2.1版本了，ttsfrd是在哪里发布新的版本？主要是和最新的autolabel1.1.8以及1.0.9resource里ttsfrd的resource不匹配啊，每次都得自己修改成上个版本的resource路径

RH531273646

运行过程中，不断打印“Load pinyin_en_mix_dict failed”

7

运行过程中，不断打印“Load pinyin_en_mix_dict failed”。虽然能正常输出音频，但不知道这条log是否表明运行有问题？是不是我配置还缺了啥？ 直接用的是SambertHifigan语音合成-中文-多人预训练-16k模型。 主角本如下： #!/bin/bash # SambertHifigan语音合成-中文-多人预训练-16k # git clone -b pretrain http://www.modelscope.cn/speech_tts/speech_sambert-hifigan_tts_zh-cn_multisp_pretrain_16k.git # speaker_list: 'F7,F74,FBYN,FRXL,M7,xiaoyu'} all except M7 are female res_zip=../funtts/speech_sambert-hifigan_tts_zh-cn_multisp_pretrain_16k/resource.zip am_ckpt=../funtts/speech_sambert-hifigan_tts_zh-cn_multisp_pretrain_16k/basemodel_16k/sambert/ckpt/checkpoint_980000.pth voc_ckpt=../funtts/speech_sambert-hifigan_tts_zh-cn_multisp_pretrain_16k/basemodel_16k/hifigan/ckpt/checkpoint_2000000.pth spk=xiaoyu outdir=out_$spk [ -d $outdir...

airstillblue

from modelscope.outputs import OutputKeys from modelscope.pipelines import pipeline from modelscope.utils.constant import Tasks from modelscope.models.audio.tts import SambertHifigan import os import time model_dir = "/home/gpu/KAN-TTS/pretrain_work_dir/" # the path of my trained speech...

hzl256

Traceback (most recent call last): File "kantts/bin/train_sambert.py", line 224, in train trainer.train() File "/home/l/project/KAN-TTS_/kantts/train/trainer.py", line 210, in train self.train_epoch() File "/home/l/project/KAN-TTS_/kantts/train/trainer.py", line 223, in train_epoch self.check_eval_interval() File "/home/l/project/KAN-TTS_/kantts/train/trainer.py", line 200,...

laishujie

CUDA_VISIBLE_DEVICES=0 python kantts/bin/train_hifigan.py --model_config speech_sambert-hifigan_tts_zh-cn_multisp_pretrain_16k/basemodel_16k/hifigan/config.yaml --resume_path speech_sambert-hifigan_tts_zh-cn_multisp_pretrain_16k/basemodel_16k/hifigan/ckpt/checkpoint_2000000.pth --root_dir newtest/training_stage/SSB0009_feats --stage_dir newtest/training_stage/SSB0009_hifigan_ckpt cuFFT error: CUFFT_INTERNAL_ERROR Traceback (most recent call last): File "kantts/bin/train_hifigan.py", line 171, in train trainer.train() File "/home/dufei/git/KAN-TTS/kantts/train/trainer.py", line 199,...

dushaofan

Infering is so slow.

2

Great work! I wanna know how to speed up infering, need help!

SaltedSlark