[Feature Request]: Batch processing different models / sampling methods / images

[cyberius0]

Is there an existing issue for this?

• [X] I have searched the existing issues and checked the recent builds/commits

What would your feature do ?

Hello,

I couldn't find any specific information on this, so that's why I ask. Is there a possibility to create "batch jobs", e.g. choose a txt2image prompt, model, sampler and "save this job", then choose another sampler, save this other job, maybe then do some img2img jobs and in the end start them all, so they will be worked on one by one?

Proposed workflow

Possibility to "save" the current workflow / settings and add it to a batch, where all jobs can be started later.

Additional information

No response

[kleer001]

Depending on how different you need your batches (and I do think scheduling batches would be powerful and cool) you can set them up in a single "Generate" pass with the powerful X/Y(/Z) plot Script that's built in.

Using the variable "Checkpoint Name" would let you change models. "Prompt S/R" is really awesome and if you needed you could rewrite a whole prompt if need be. Then there's "Prompt Order" too.

But that's only a possible work around.

[Jonseed]

You should be able to do this with the "prompts from file or textbox" script on the main txt2img tab. Here is some documentation about how to use it: https://github.com/AUTOMATIC1111/stable-diffusion-webui/issues/1739#issuecomment-1317022160

[mezotaken]

https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Features#prompts-from-file-or-textbox Updated wiki on it, yep, it works just like you'd like it to work.