Frequent freezes of gpt-3.5-turbo-0125 - I don't wait for the response

I switched to the gpt-3.5-turbo-0125 model. And I started encountering hanging errors. I don’t receive any errors from the API, it just hangs upon request. However, if I check the Assistant logs at the same time, everything is in order there, no errors occurred, and the response is already in the logs. But I don’t receive it.

I experienced the same issue in the Playground Assistant. The RUN button has not been activated for 10 minutes and it is stuck at the “Add a message” stage.