The dataloader, train_dataloader, does not have many workers which may be a bottleneck

[heavy niche]

I'm seeing this message when I run TI training. The full message:
The dataloader, train_dataloader, does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` (try 10 which is the number of cpus on this machine) in the `DataLoader` init to improve performance.

I'm on an M1 Max MacBook Pro. I'm not sure how to make the change it's recommending. Would this actually improve training performance on my machine?

I'm also curious if it's possible to parallelize GPU work (if that's not happening already). When I look in Activity Monitor, I can tell if multiple CPU cores are being used if a process is using >100% CPU. But I never see anything exceed 100% GPU, even though the GPU in this machine has 24 cores. This is probably just some fundamental thing I'm not understanding, but it makes me wonder if my hardware is underutilized by InvokeAI.

[icy summit]

Hi there, this is talking about a few variables you can change in the code to increase speed of training. However, I would not do that unless you believe you have the available compute resources to handle that. https://towardsdatascience.com/how-to-fine-tune-stable-diffusion-using-textual-inversion-b995d7ecc095

This link will walk you through what to change if you want to go that direction