The nvOC automation files are not designed for that, and there's a lot to do without trying to add that complexity.

That beings said, some mining software does allow you to do that. For example, Claymore lets you specify which GPUs you want it to use. You can fire up one instance of Claymore with 7 cards, and then another instance with 6 cards, each on a different algo that Claymore supports.

I don't know which miners support picking/choosing GPUs, just that Claymore does.

That said... why do you want to do this? It doesn't make a lost of sense to me.

When there's a hot Equihash coin, like Zclassic is at the moment, usually all the NVidia GPUs seem to profit more on it than on, say, ETH, so you might as well point them all at it...

Meanwhile, if you're looking to split up your mining so that you can accumulate two different coins, you can just mine one day per coin, and keep switching back and forth. You'll still accumulate... and the rig will probably be more efficient running a single miner process, and managing/supporting it will be easier to troubleshoot & etc. than if you tired to run two different mining processes.

(Personally, I've separated all my rigs into AMD-only/1060-only/1070-only rigs, so as not to mix GPUs specifically in order to avoid having to split the GPUs.)