'Model is currently loading" problem for microsoft/speecht5_tts interference API

I’m trying to use the interference API for microsoft/speecht5_tts and Clinical-AI-Apollo/Medical-NER both showing the same message
b'{"error":"Model microsoft/speecht5_tts is currently loading","estimated_time":23.41907501220703}'
and
{'error': 'Model Clinical-AI-Apollo/Medical-NER is currently loading', 'estimated_time': 29.423261642456055}

It’s still the same after waiting for the estimated time and trying again.
I’ve used
output = query({ "inputs": "Hi", "options":{ "wait_for_model": True} })
to wait for model and it shows b'Internal Server Error'