Quantcast
Channel: Service topics
Viewing all articles
Browse latest Browse all 62482

Copilot Context Issue in Fabric Notebook

$
0
0

Hi all,

 

In Fabric, I'm trying to use Copilot in a Notebook. I have a DataFrame df that I created by reading a table from my lakehouse. When I try to run the following code asking Copilot to summarize the data in df:

 

%describe df

 

 

 I receive the following error:

 

RuntimeError: StructuredContext.optimize() was called with a mandatory context that was too large to fit in the max_prompt_tokens limit. The mandatory context was 3626 tokens, but the limit was 3346 tokens.

 

 

It's clear an input exceeds the maximum token length. I have nothing else loaded into memory other than df, which has a shape of (137923, 13) though the error still occurs if I try with only 1,000 rows instead. I am not aware of a way to control the context that Copilot includes before passing to gpt-35-turbo-0125. Other magic commands such as %%chat and %%code work normally without issue.

 

Has anyone encountered this or have any tips or things to try? Also, if there's a better place to post this, please let me know. Thanks!


Viewing all articles
Browse latest Browse all 62482

Trending Articles



<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>