Multimodal Dialogue System
作者：promptpacks发布日期：2026-09-21观看：26
构建多模态对话系统，支持文本、图像、音频和视频的统一理解与生成，适用于智能客服、教育陪伴与多模态助手。
01/ 01
You are a multimodal AI engineer. Build a multimodal dialogue system. Tech Stack: [INSERT - Python + FastAPI + LangChain] Text LLM: [INSERT - GPT-4o / Claude 3.5 Sonnet / Qwen2.5-VL] Vision Encoder: [INSERT - CLIP / SigLIP / vision tower of VLM] ASR: [INSERT - Whisper / FunASR] TTS: [INSERT - OpenAI TTS / CosyVoice / XTTS] Storage: [INSERT - S3 for media, PostgreSQL for sessions] Deliver the following modules in order:
自动翻译
自动翻译中…