Share via

Azure AI Foundry text prompt flow endpoint issue

Farah Diana Masri 40 Reputation points
2025-01-13T01:16:14.09+00:00

Hi,

I have issue regarding text prompt flow. I already deploy a text prompt flow in Azure AI Foundry. This prompt flow is working fine before, but now the prompt flow cannot be used. I don't know why, because I did not modify the prompt flow, but it seems like the error is coming from prompt flow. Please assist me on this matter. Below are the error logs in the text prompt flow endpoint and error show in test tab.

Thank you for helping!

Error in test tab:
Screenshot 2025-01-13 085406

Azure AI Search
Azure AI Search
An Azure search service with built-in artificial intelligence capabilities that enrich information to help identify and explore relevant content at scale.
1,142 questions
Azure AI services
Azure AI services
A group of Azure services, SDKs, and APIs designed to make apps more intelligent, engaging, and discoverable.
3,042 questions

1 answer

Sort by: Most helpful
Most helpful Newest Oldest
  1. Shikha Ghildiyal 2,170 Reputation points Microsoft Employee
    2025-01-13T04:44:34.9866667+00:00

    Hi Farah Diana Masri,

    Thanks for reaching out to Microsoft Q&A.

    This error usually comes from issues in Azure PromptFlow’s tracing module, where None values in token usage data cause the TypeError. A quick workaround is to disable tracing by adding this environment variable to your flow.dag.yaml file:

    YAMLCopy

    

environment_variables: 

	PF_DISABLE_TRACING: true

    Save and redeploy your service. Disabling tracing will bypass the error, allowing a smooth deployment if tracing data isn’t essential. Troubleshooting steps can be found here- Deploying workflows in Azure PromptFlow can run into unexpected issues for a deeper dive into resolving this error in Azure PromptFlow.

    Please do not forget to "Accept the answer” and “up-vote” wherever the information provided helps you, this can be beneficial to other community members.

    0 comments

Your answer

Answers can be marked as Accepted Answers by the question author, which helps users to know the answer solved the author's problem.