请求新增ocr功能 #1344
aaronliDev6666
started this conversation in
Ideas
请求新增ocr功能
#1344
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
发送图片时自动 OCR 提取文字,兼容纯文本模型DeepSeek API,在 RikkaHub 里传图片会直接报错:
因为 DeepSeek 是纯文本模型,不吃多模态格式。图片发不出去,挺难受的
能不能在 RikkaHub 前端加一个 本地 OCR 预处理?这样 DeepSeek 等纯文本模型也能「看图」了~
btw,rikka真的很成功🫡
Beta Was this translation helpful? Give feedback.
All reactions