RVC-Project/Retrieval-based-Voice-Conversion-WebUI

Getting weird accents

Open

#361 创建于 2023年5月27日

在 GitHub 查看
 (2 评论) (0 反应) (0 负责人)Python (18,427 star) (2,849 fork)batch import
documentationhelp wanted

描述

The voice of the audio files processed with my model sounds like me, but with strange accents. Can only certain languages be spoken/sung cleanly? Or did something go wrong during training or something?

贡献者指南

技术栈
pythonpytorch
领域
machine learningai
议题类型
research
难度面向新贡献者的预计实现难度，1 表示很小改动，5 表示专家级工作。
4
预计时间有经验贡献者完成调查、实现、测试并准备 pull request 的粗略时间范围。
1-3 hours
活动状态议题当前的可参与程度：新鲜、活跃、陈旧、阻塞或等待维护者输入。
fresh
清晰度议题是否清楚说明期望改动、验收标准和下一步。
unclear
前置要求
basic machine learning conceptsPython
新手友好度1-100 的估计分数，表示该议题对首次贡献者的友好程度。
20
研究方向
First, check the training data used: was it multilingual or only one language? Look at the model's training script (e.g., train.py) and see if any language specific preprocessing was applied. Also review the inference pipeline to ensure no accidental accent modification. The issue #361 may need more details from the user about which model and training configuration they used.