from transformers import pipeline
pipe = pipeline("automatic-speech-recognition", model="stt-thai",return_timestamps=True)
result=pipe("thai.wav", generate_kwargs={"task": "transcribe","language":"th"})["text"]
print (result)