How to serving multi-gpu inference？

Question

Alone-wl opened this issue a year ago · 1 comments

Answer 1 · 2023-12-21T02:31:10.000Z

mpirun -np ${num_of_GPU} python xxx.py