Closed lixc123 closed 4 days ago
I'm facing the same issue as the previous user. I'm using Azure GPT4O Mini to create workflows, and even with security risk control set to the lowest level, there's still a chance that Azure's risk control will be triggered, interrupting the workflow. I'm also curious if there's a way to make the workflow automatically detect and switch models when a risk control is triggered, so the workflow can continue running. Have you encountered a similar situation? If so, let's discuss how to address this issue.
Hi, @lixc123. I'm helping the Dify team manage their backlog and am marking this issue as stale.
Your issue raises important concerns about Azure's security risk control potentially disrupting workflows with the GPT-4o mini model, even at the lowest security settings. User JR-DDATA has also shared similar experiences and inquired about the possibility of automatically switching models when a risk control is triggered. It seems both of you are interested in discussing potential solutions to this problem.
Could you please let us know if this issue is still relevant to the latest version of the Dify repository? If it is, feel free to comment here to keep the discussion going. Otherwise, you can close the issue yourself, or it will be automatically closed in 7 days. Thank you!
Self Checks
1. Is this request related to a challenge you're experiencing? Tell me about your story.
I use azure gpt4o mini to make workflows. The security risk control has been adjusted to the lowest level. However, there is still a chance that azure risk control will be triggered, which will cause the workflow to be interrupted in the middle. What I want to ask is whether the workflow can be continued without interruption, so that if can judge and switch the model to respond correctly.
2. Additional context or comments
No response
3. Can you help us with this feature?