I can run "A1111" just fine, but this fork - not so much

[space-hamster0442]

Hello.
I can run "A1111" just fine, but this fork - not so much...

windows 11
16gb ram, 2060super, 6gb vram (laptop version)
Python version: 3.10.11, Git: 2.40.0, Cuda: 12.1.1_531.14

I'm getting "CUDA Out of memory" and can't generate a single image. My settings for running "A1111" were "set COMMANDLINE_ARGS= --medvram --precision full --no-half --xformers", "set PYTORCH_CUDA_ALLOC_CONF=garbage_collection_threshold:0.6,max_split_size_mb:512".

Maybe you can help me with specific settings to run this fork, I really want to try it, I'm getting so much errors on "A1111" with the latest update(s).

edit: sorry, if this question was asked before, I'm new here

[space-hamster0442]

Sorry, did a "clean" reinstall, now images are generating... But I still don't understand how to add command line params here...

[gsgoldma]

dont think you need to. i have 6 gigs and can make an 512 bs 512 image in 3-4 seconds using only 30-50% of my vram

[TingTingin]

the command line params are moved to the ui so you can edit them from the ui directly

[space-hamster0442]

Thank you for your help, looks like I have a new problem...
When I try to use "Interrogate CLIP" in img2img I get this:

...in txt2img too, when I try to use hires.fix
edit: small typos

[vladmandic]

@space-hamster0442 that has nothing to do with original issue and it seems like download of model is not working

[space-hamster0442]

Sorry, just thought about it but decided to write about it here anyway, you can move it...
Thought that it still can be linked with the original topic.