What's the speed of generate instruction?

[LeopoldACC]

Hi, @percyliang , thanks for your sharing! But when I try to regenerate instruction by script generate_instruction.py, the cost of time seems so long, even if I reduce the number of instruction from 20 to 8 in each prompt and batch_size from 5 to 1. So, what's the speed of your data generation when setting is model:text-davinci-003, number of instruction:20,batch_size:5?

Another problem is that I can not get response when setting is the same with you, the process is stop at the