[Bug]: Error getting prompts to prompts.zip

[michalmar]

Do you need to file an issue?

• [X] I have searched the existing issues and this bug is not already filed.
• [X] My model is hosted on OpenAI or Azure. If not, please look at the "model providers" issue and don't file a new one here.
• [X] I believe this is a legitimate bug, not just a question. If this is a question, please use the Discussions area.

Describe the bug

I am following 2-Adavnced-getting-started.ipynb and generate_prompts(storage_name=storage_name, limit=1, zip_file_name="prompts.zip") fails getting prompt from APIM request:

Error:

Error generating prompts for data in 'my-container'. Please try a lower limit.

Steps to reproduce

I am following

Error:

Error generating prompts for data in 'my-container'. Please try a lower limit.

Expected Behavior

I would download the prompts to local folder.

GraphRAG Config Used

# Paste your config here

Logs and screenshots

No response

Additional Information

• GraphRAG Version:
• Operating System:
• Python Version:
• Related Issues:

[jgbradley1]

Hi @michalmark, I think your question is related to the solution accelerator (and not the graphrag library specifically). I have seen this issue before.

In our testing, we believe it is due to the fact that the APIM service that gets deployed by default is set to use a Developer tier sku. The Developer tier has no SLA and therefore, intermittent connection issues between APIM and the client that initiated the API request (you) can occur.

When/if you deploy the solution accelerator and modify the APIM Tier to use the newer StandardV2 sku, this issue does not show up for me. Just be aware that StandardV2 cost more and is the reason why we have the default set to Developer.

If you want to use the prompt tuning API endpoint with an APIM deployment based on the Developer tier, I usually have success by rerunning the notebook cell that generates the prompts a few times.

[michalmar]

Yes, in my case its Developer and the issue is persistent, a.k.a. even after multiple tries I am getting this error.

However, I was able to use workaround when i manually created the prompt files with graphrag.index --init command.