-
|
Hi, I am trying to support images for my local GPT assistant. Will Connery Action be able to support outputting images? It doesn't appear so if I am reading Thanks! |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments
-
|
Hi @mhaowork, you're right - currently, only string outputs are supported. As a workaround, you may convert the image to base64 and return it as a string. Then, on a client, convert it back to the image. Would it work in your case? |
Beta Was this translation helpful? Give feedback.
-
|
Thanks for the reply! Sounds good!
…On Tue, Apr 23, 2024 at 12:16 AM Volodymyr Machula ***@***.***> wrote:
Hi @mhaowork <https://github.com/mhaowork>, you're right - currently,
only string outputs are supported. As a workaround, you may convert the
image to base64 and return it as a string. Then, on a client, convert it
back to the image. Would it work in your case?
—
Reply to this email directly, view it on GitHub
<#117 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ACCMXBO5XAU3BNAPVBBOG33Y6YDFTAVCNFSM6AAAAABGTYFGWGVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TCOJXGM2DK>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
Beta Was this translation helpful? Give feedback.
Hi @mhaowork, you're right - currently, only string outputs are supported. As a workaround, you may convert the image to base64 and return it as a string. Then, on a client, convert it back to the image. Would it work in your case?