跳转到内容
tansr 资料站

AppBundlePlatformModels

授权媒体模型集(bundle platformModels 段):顺序 = 平台下发的稳定序,顺位第一即工具 / 网关未指定 model 时的生效模型;空数组 = 平台未配置(如实呈现,恒不猜名)。语音识别与语音合成是两个独立面, 键名用全词 speechToText / textToSpeech(wire 面对应 /t1/asr、/t1/tts)。

属性

Section titled “属性”

imageGen

Section titled “imageGen”
imageGen: ImageModelDescriptor[];

speechToText

Section titled “speechToText”
speechToText: SpeechToTextModelDescriptor[];

textToSpeech

Section titled “textToSpeech”
textToSpeech: TextToSpeechModelDescriptor[];

videoGen

Section titled “videoGen”
videoGen: VideoModelDescriptor[];