Closed JackTesla closed 1 year ago
Please use torchrun as mentioned in INSTRUCTION.md.
torchrun --nproc_per_node=<num-of-gpus-to-use> --master_port=23456 main.py \ --cfg <config-file> --data-path <imagenet-path> [--batch-size <batch-size-per-gpu> --output <output-directory> --tag <job-tag>]
Actually, I used torchrun too. The error still exists.
This is an error caused by main.py which I've commited.
Yes, you are right. Resolved.
When I run the following command :
It outputs this: