The model cannot be trained with multiple cards on a single machine

[langge52]

Dear author, hello

Thank you very much for sharing these codes.

The problem I am currently facing is that I am unable to train with multiple cards on a single machine. Due to the abandonment of torch. distributed. launch, I have attempted CUDA_ VISIBLE_ DIVICES=0,1,2,3 Python - m torch. distributed. run -- nnodes 1-- nproc_ Per_ Node 4 train.py -- config configs/demo. yaml; Torchrun train.py -- config configs/demo.yaml and other training commands cannot be trained, and there is no relevant log information output. Therefore, I would like to ask you for advice on how to solve this problem. Thank you very much and look forward to your reply. Thank you again.