Making an Ollama Model read dict/json data properly

penguin_pam

I am using Big Query to pull SQL data from a certain database. I am having issues making an Ollama model, particularly the llama3.3 70B, making it read the SQL query data properly. It only just reiterates the given prompt and asks for more information. I convert the SQL data into a dictionary, and convert the dictionary to json data through the .jsondump(). If anyone has any suggestions so that I can make it read it properly. My main goal is for...

Originally asked by Bryce Heerink on Stack Exchange (stackoverflow) (CC BY-SA 4.0).

uphilltech

Great question — answering as the UphillTech team:

Use json.dumps instead of json.dump to assign the data to a variable, see distinction: https://www.geeksforgeeks.org/python/python-difference-between-json-dump-and-json-dumps/

Full answer by Sébastien Guarnay: https://stackoverflow.stackexchange.com/a/79854642 (CC BY-SA 4.0).

Need it done for your business? UphillTech.