Skip to content

Weired Generation #5

Description

@John4509

I used single_inference like
python single_inference.py \ --content_to_synthesize "今天天氣真好,我們來烤肉吧" \ --speaker_prompt_text_transcription "在密碼學中,加密是將明文資訊改變為難以讀取的密文內容,使之不可讀的方法。" \ --speaker_prompt_audio_path "./data/example.wav" \ --model_path "./models"
However, it only generate "(no voice) 我們來烤肉吧"

when I tried
python single_inference.py \ --content_to_synthesize "我們來烤肉吧" \ --speaker_prompt_text_transcription "在密碼學中,加密是將明文資訊改變為難以讀取的密文內容,使之不可讀的方法。" \ --speaker_prompt_audio_path "./data/example.wav" \ --model_path "./models"
it only generate "(no voice)"

when I tried
python single_inference.py \ --content_to_synthesize "。我們來烤肉吧" \ --speaker_prompt_text_transcription "在密碼學中,加密是將明文資訊改變為難以讀取的密文內容,使之不可讀的方法。" \ --speaker_prompt_audio_path "./data/example.wav" \ --model_path "./models"
it only generate "(hallucination)我們來烤肉吧"

How can i solve this?

Metadata

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

    Issue actions