To run a batch sequential, use the node 'Image Batch to Image List' For Latent there is no such node, but you can get around that with the Node pack 'Inspire' From your loader connect latent output to 'Rebatch Latents'. Set the batch size to 1. Then connect the output from 'Rebatch Latents' to 'Random Generator for List (Inspire)', and also to the ksampler latent input. On your ksampler set seed to an input instead of widget. From 'Random Generator for Last' go to the seed input. This is more economical and faster than rerunning the entire workflow over and over. Especially if it's further down the line.
ControlNEt with LORA is the same workflow, just put the LORA box(es) after loading hte module. Regarding animateDiff, I am working in a tutorial but it is very straightforward to use.
To run a batch sequential, use the node 'Image Batch to Image List'
For Latent there is no such node, but you can get around that with the Node pack 'Inspire'
From your loader connect latent output to 'Rebatch Latents'. Set the batch size to 1. Then connect the output from 'Rebatch Latents' to 'Random Generator for List (Inspire)', and also to the ksampler latent input.
On your ksampler set seed to an input instead of widget. From 'Random Generator for Last' go to the seed input.
This is more economical and faster than rerunning the entire workflow over and over. Especially if it's further down the line.
thanks for the tip
Thanks for showing t he way much faster way 😁 Thanks, this was helpful.
How to implement controlNet into LoRA workflow? Also with animateDiff (text2vid, etc)
ControlNEt with LORA is the same workflow, just put the LORA box(es) after loading hte module. Regarding animateDiff, I am working in a tutorial but it is very straightforward to use.