Auto queue or batch execute the whole workflow, while the for loops execute only a part of the workflow - while you could probably rig up a batch / auto queue logic that executes fundamentally in the same way, it’s much simpler to use loops if your goal is to reiterate only a portion of a workflow
nowadays I'm working on my custom runpod servers mostly, but anything that's happening here is fairly straightforward in terms of memory requirements - maybe the iterative upscaler at very high resolutions will run out of memory sooner or later, but the other workflows are rather lightweight
Really great tip! Thanks for sharing Andrea
Baioni regala emozioni 🖤🖤
Always nice to learn AI from john lennon :) hehe Thanks a lot man!
thanks for introducing this Loops node. Do you know if replaces the use of Queue's Batch Count in some cases?
Auto queue or batch execute the whole workflow, while the for loops execute only a part of the workflow - while you could probably rig up a batch / auto queue logic that executes fundamentally in the same way, it’s much simpler to use loops if your goal is to reiterate only a portion of a workflow
Yeeeeah loops
Will this work on a regular rig or do you need some abomination to run it?
nowadays I'm working on my custom runpod servers mostly, but anything that's happening here is fairly straightforward in terms of memory requirements - maybe the iterative upscaler at very high resolutions will run out of memory sooner or later, but the other workflows are rather lightweight