-
The tone result of speech conversion is very different from the voice of the target speaker. Is there any direction that can be optimized?
Here is my training code:
```
import os
from trainer …
-
請問我在輸入 `python demo_toolbox.py -d C:\aishell3_data` 後卻沒有像視頻一樣預先載入訓練好的Dataset是哪邊出狀況了嗎?
(我的檔案存放路徑: `C:\aishell3_data\aishell3`)
![Inked1630119737056_LI](https://user-images.githubusercontent.com/55…
-
如题,tts 中,我想给 aishell3 数据里额外添加一些数据来进行训练(采样率相同),对于 am 和 voc,请问我除了需要 '文本内容' 和 '音频数据' 外,还需要其他东西吗?我看到其他 issue 里面说,直接给 aishell3 的数据里面加一个 speaker_id 即可,那除此之外的步骤能大概描述一下吗?非常感谢!
-
运行 run_gen_en_new.sh 出现如下问题:
```
root@container-c6d2118d3c-6824b4ba:~/xx/PaddleSpeech-r1.1.0/examples/ernie_sat# ./run_gen_en_new.sh
new_str is We are trying to establish a date.I enjoy my life…
-
**Describe the bug**
Always the same speaker output in fastspeech2 aishell3 voice conversion
**To Reproduce**
1. [PaddleSpeech语音克隆](https://aistudio.baidu.com/aistudio/projectdetail/3475121?chann…
-
Introduction:
- https://github.com/PaddlePaddle/ERNIE/tree/repro/ernie-sat
PR:
- https://github.com/PaddlePaddle/PaddleSpeech/pull/2052
- https://github.com/PaddlePaddle/PaddleSpeech/pull/2117
…
-
目录examples/aishell3/vits-vc训练的模型为啥效果不好?V100单卡GPU训练了3天,mel_loss下不来一直在26左右徘徊。可以给出你们paddlespeech训练好的模型文件瞅瞅么?
日志示例:
INFO 2022-09-20 10:58:28,841 trainer.py:167] iter: 74635/350000, Rank: 0, real_loss: …
-
实现单模型TTS中英文发音初步尝试了3种方案:单发音人方式、多发音人语料混合方式、finetune方式。
示例句子:
```text
大家好,我是parrot虚拟老师,我们来读一首诗,我与春风皆过客, I and the spring breeze are passing by,你携秋水揽星河, you take the autumn water to take the galaxy。
…
-
准备 aishell3 时 94 秒一个人。
conda python 3.9 cuda 10.2 torch 1.9.0 vision 0.10.0
![image](https://user-images.githubusercontent.com/65227354/179400835-43cd51cc-ec0a-4601-8fe5-91df315999be.png)
-
when i use the config "python synthesize.py --text "大家好,欢迎收听我的频道" --speaker_id 0 --restore_step 100000 --m
ode single -p config/AISHELL3/preprocess.yaml -m config/AISHELL3/model.yaml -t config/AISHE…