Find answers from the community

Updated 3 months ago

i have a qestion if i am using agents

i have a qestion if i am using agents using llama_index and the user input containes grmmer or spilling error how can i fix it or tell the agent to understand it?
L
C
4 comments
most LLMs just kind of understand πŸ˜…
@Logan M ya but am using google gemini and agents using RAG so it can only answer the Q&A on my docs but if the qestion is contain spliing error i does not auto correct it just says Sorry, I cannot answer your query.
@Logan M for exmaple:

i asked the agent this qestions what juz is surah al fatina altho the sentence is correct but i replace the word Fatihah with fatina so little spiling error and the agent respons with:
Enter a prompt (q to quit): what juz is surah al fatina Thought: I need to use a tool to help me answer the question. Action: juz_data Action Input: {'input': 'a patina'} > Pandas Instructions: ``` df.query("Surah.str.contains('a patina')") ``` > Pandas Output: Empty DataFrame Columns: [Surah, Juz] Index: [] Observation: Empty DataFrame Columns: [Surah, Juz] Index: [] Thought: The input surah name is not valid. I cannot answer the question with the provided tools. Answer: Sorry, I cannot answer your query. Sorry, I cannot answer your query.

but since that was little grammer error i want it to auto correct it-self using the docs is there a way
Add a reply
Sign up and join the conversation on Discord