Llama3.2-11B-Vision-Instruct does not remember the context when using with agents and sessions

#75
by Irfanraza - opened

Here is the detailed description of the problem.

When i send the single text prompt it works fine.
When i send the follow up text prompt it remember the context and answer it, so it works fine.
Same thing follows for text again and again.
When i send the image with text prompt it works fine and answer it.
When i send the text after sending (image with text) it forget’s the context and says i don’t remember any image which you have sent.
Does anyone have faced similar issue? or any solution to this problem?

Sign up or log in to comment