Open biancafrigerioMS opened 1 year ago
The image generation prompt (auto generated) can trigger sensitive data filters in OpenAI and expecially in Azure OpenAI. In subsequent versions I have extracted only the DallePrompt text and tune the DallePrompt generation to focus on the face/portrait. This reduces the probability of triggering AI safety filters.
For Azure OpenAI, you can request to have greater control on the AI Safety as you are working on clinical scenarios so that you can increase the tolerance.
Hi, I uploaded without errors the app and when testing the Chat it gives me the following error in each "text box" section: OpenAIapi.ChatNoFunctions failed: scrubbedSensitiveData ...
Open AI has correctly connected and also Azure Open AI.
Can you help me in understanding how I can fix it?
Thank you
Bianca