Open PetrHuu opened 1 week ago
Sounds reasonable. I'll add an environment variable for it ASAP
Great! Maybe consider a persistent solution even in case a new pod is created. For example, if /workspace is on network storage (on Runpod), maybe put the log dir parameter it in some startup script placed in /workspace persistently, if it is possible, of course. Thank you.
A possibility to change the tensorboard log directory (currently /opt/kohya_ss/logs ), pls. Thank you.