I want to create a private discord bot that creates a tl:dr for all the messages around a discussion.
I used gemma3:12b to create a tl:dr for around 380 discord messages but the result seems to be not accurate.
I am a total beginner so I am not even sure if thats the right or best model for this job. It seems to work good on just a few messages (~20).
I only want to feed text to the AI with a single prompt and get the tl:dr as result.
Should I switch to a different model?
The prompt I generated with chatgpt (because I have no clue about good prompts) that gets feeded to the AI is:
You are a professional Discord summarization assistant. Your task: - Summarize the messages of a Discord channel. - Identify discussions. - Identify different opinions. - Attribute statements to the respective people. - Ignore small talk as much as possible. - Highlight decisions and outcomes. - Respond in German. [Length prompt] IMPORTANT: If different people have expressed different viewpoints, create a section: ## Positions and list the respective stances. If no discussion took place, omit this section. Messages: [List of messages] [Length promt] gets replaced with something like:
Medium-length summary. Approx. 8–15 bullet points. Mention key topics and outcomes. [List of messages] do have the format of "user: message \n".
Is it alright to feed the AI all the messages at once?
[link] [comments]