Spaces:
Runtime error
Report History bug
Nice bot! I love it!
There is a bug with the store_report() function, I think that is called only after clicking submit_btn. So you can't see history until you execute something.
And a recommendation could be to put a button to stop the execution.
Keep working on it! its a really nice project
Thanks for your feedback!! I will make sure to address the issues as soon as I find some time after completing my dissertation. Your input is greatly appreciated!
Nice bot! I love it!
There is a bug with the store_report() function, I think that is called only after clicking submit_btn. So you can't see history until you execute something.
And a recommendation could be to put a button to stop the execution.
Keep working on it! its a really nice project
Hi GianJSX, the issue has been fixed! (Note that this is only a demo and everybody shares the same api key and memory so far. So please don't input something private. :)
Feel free to test out the system and let me know if you encounter any further issues or have any feedback.
@zej97 I'm working on custom prompts. What do you think about adding something like this, maybe like an advanced settings
I can share the code changes, but it's my first time using Gradio. Maybe you can review the code and suggest if you consider adding it.
Without commitments, it is only a proposal that is useful for me.
That is so great! Feel free to pull a request on github.
Also, there is a problem with long responses that end incomplete.
In my case, I replaced the fast_llm_model from 3.5-turbo to 3.5-turbo-16k, and the token limit from 8000 to 16000.
But I think that could be nice to have a "continue writing" button, like chatgpt.
I think that the "continue writing" button will need some kind of summarization. Maybe some intermediate agent to summarize