Closed q894257739 closed 2 months ago
This issue is stale because it has been open 60 days with no activity. Remove stale label or comment or this will be closed in 10 days.
This issue was closed because it has been stalled for 10 days with no activity.
I encountered two problems when using gpt-4o-2024-05-13: 1.When I use 4o to organize the document content and output, when the document content is greater than 1200token, it will not respond until timeout (I am using AsyncAzureOpenAI under openai, a model deployed on Azure)
2.When I use 4o to organize the contents of the document and output, I use json_model to output json, normally I can output normal content, but after 20 requests, there are a few times to return a strange floating point number (1.0,2.0,3.2...). I make sure that my document and prompt words do not contain any of these numbers, and any intention to output these numbers, the normal response time is 3 seconds to 7 seconds, but once the number is returned, the response time is only about 2s, the same prompt words,gpt3.5 will not appear such an exception, I want to know is the reason for the model, or said Azure The reason is because the model is deployed on Azure, and the subsequent workaround