sophgo / tpu-mlir

Machine learning compiler based on MLIR for Sophgo TPU.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

How could I have run the executable on the my CPU/your TPU?

Z-KN opened this issue · comments

commented

This is a very nice project, but I still have some confusion in the first place.

I saw there are two target hardware backends, BM168x, and CV18xx. Users (like me) don't have such hardware, but I could run the detect_yolov5.py script without errors following the README steps and get the dog_int8.jpg inference results. But how? Is the .bmodel file the actual binary executable on your TPU? If yes, how could I run the .bmodel on my CPU? What actually happened here?