Multiple background server for multi GPUs?

[nvdamien]

Describe the solution you'd like The CLI is checking if a server is already running in the background here: https://github.com/Comfy-Org/comfy-cli/blob/main/comfy_cli/command/launch.py#L143

But I want to run one server for each GPU (on different port). Like, I have 1 node (H100) with 8 GPUs, I would run 8 servers in the background with different port (8188, 8288, 8388, 8488, 8588, 8688, 8788, 8888) and the extra GPU arg to select the GPU I want to run.

Describe alternatives you've considered Can we add a flag to "force" to be able to run multiple servers? Or remove the check completely, and add some arg to specify what port we want to stop when we run "comfy stop"

I can do a MR if you tell me what direction we want to take here.

Edit: Started to work on that. I have something working. Testing more.