Hi, I need guidance on how to best instruct the Large Language Model (LLLM) in a data summarization prompt to format the output in a specific way, like bullet points. Currently, it only outputs hyphens ('-') for bullet points and asterisks ('*') for new lines. Is it possible to influence the format somehow through the prompt?
Thanks.
Hi, from experience, it just a try and error-process. If after several tries it does not work, you can do 2 things:
1. Call a webhook to format the output in the way you want
2. You can use the SUBSTITUTE system function: https://cloud.google.com/dialogflow/cx/docs/reference/system-functions#func-substitute
Best,
Xavi
User | Count |
---|---|
2 | |
1 | |
1 | |
1 | |
1 |