遇见数据集

rumik-ai/emilia-s2s-mimi-q8-named

收藏
Hugging Face2026-05-26 更新2026-06-14 收录
官方服务:

资源简介:

Emilia S2S Mimi Q8 Named Speakers是一个基于TinyAya codec-tokenized的文本到语音(TTS)数据集,从指定路径的parquet文件构建。数据行包含Mimi codec-only示例,训练文本字段已预置面向用户的说话人前缀(例如Ira: 转录文本...)。说话人映射关系为:SP_SP010对应Ira:,SP_SP061对应Aisha:,SP_SP073对应Siya:,SP_SP079对应Zoya:。该数据集的参考条件被禁用,不输出任何ref_*字段。

Emilia S2S Mimi Q8 Named Speakers is a TinyAya codec-tokenized TTS dataset built from a specific parquet file path. Rows are Mimi codec-only examples, and the training text field already includes user-facing speaker prefixes (e.g., Ira: transcript...). Speaker mapping is provided: SP_SP010 -> Ira:, SP_SP061 -> Aisha:, SP_SP073 -> Siya:, SP_SP079 -> Zoya:. Reference conditioning is disabled for this artifact, and no ref_* fields are emitted.

提供机构:
rumik-ai
二维码
社区交流群
二维码
科研交流群
商业服务