How to make the model work on the new M4F chip using external extension SPI Flash and SPI PSRAM instead of my chip's small space

I have run my voice recognition model on the development board of my ARDUINO Nano RP2040, which is very good. However, now I need three such boards to complete my creative project, but the price of this development board is not friendly. I want to use the existing M4F chip with external SPI interface Flash and SPI interface PSRAM to implement it, but I have encountered difficulties. I don’t know how the external Flash and PSRAM communicate with my model to make it work. Can you answer this for me? thanks