Skip to content

Batching API v2/batch #216

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
3 tasks
zhzLuke96 opened this issue Mar 26, 2025 · 0 comments
Open
3 tasks

Batching API v2/batch #216

zhzLuke96 opened this issue Mar 26, 2025 · 0 comments
Labels
performance Performance issues Story Next iteration summary and TODO list

Comments

@zhzLuke96
Copy link
Member

zhzLuke96 commented Mar 26, 2025

需求:支持多说话人在同一个 batching 中

目前可以batch推理,但是都是基于 同context多输入=>单输出 的形式,无法支持 不同context多输入=>多输出
计划增加一个接口 v2/batch 用于 多输入=>多输出 推理
并且,ssml推理也应该采用 batch 形式推理加速

同时这个修改应该和 #209 相关,可以同时引入并一起测试

TODOs

  • refactor batching pipeline, support m2m batching
  • v2/batch api
  • ssml based on m2m batch

参数大概类似这样:

class Segment(BaseModel):
    text: str

    prompt: Optional[str] = None
    spk: Optional[SpeakerConfig] = None

    adjuct: Optional[AdjustConfig] = None
    enhance: Optional[EnhancerConfig] = None


class BatchParams(BaseModel):
    segments: list[Segment]

    encoder: Optional[EncoderConfig] = None
    infer: Optional[InferConfig] = None
    tn: Optional[TNConfig] = None
    tts: TTSConfig = Field(default_factory=TTSConfig)
@zhzLuke96 zhzLuke96 added Story Next iteration summary and TODO list performance Performance issues labels Mar 26, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
performance Performance issues Story Next iteration summary and TODO list
Projects
None yet
Development

No branches or pull requests

1 participant