1.5B-Long 模型倾向于生成人声？ #61

Open

opened

测试中发现挺多音频中都包含含糊不清的人声，如何约束更好的生成纯音乐？

Metadata

Assignees

No one assigned

Labels

No labels

No labels

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests