your video's are not very structured. You're making it up on the fly it seems. If you write a script beforehand and stick to it, your videos would have more views.
im having real trouble getting this working for whatever reason i can seem to add the sampler node could you make a step by step instalation tutorial im sure it would help alot of people please tag me in it if you do
What the AI spills out is the most generic and effortless thing I have read in a long time. Sorry but this aint it. It elliminates the entire fun of making up your own story and your own path. This is just sad tbh.
you could probably do that with one node. xD INPUT: "can i use imagemagick to convert an image to desaturated embossed grayscale with high white balance?"" OUTPUT: "Certainly! You can combine the emboss effect with desaturation and adjust the white balance to create normal maps with high white balance using ImageMagick. [...]"
Looks great but so far, unable to install the node. Fresh portable comfyUI with manager. Progressbar is the most likely issue, but can't get that fixed up. ~ thanks for your help via github when you get a chance, would really love to try that node of yours. Cheers ~
Getting alot of milage from DSIN normal map. - Also when it works the calculate normals from lighting in IC Light seems to work well with controlnet, even if doesn't really work as a true normal map when generated from simple gradient fills.
Absolutely Amazing! Audio to Audio would be nice; really curious to try that out! Is somebody aware if finetunes or loras are currently already made for this model? How complicated is it and where to find them? Hugging face?
Yeah, still working on audio2audio, but it is coming along. As far as finetunes and loras I know civit will eventually add in an audio models section. They are working hard on their UI. Haven't looked on HF, but you could search for "stable audio open" and see if anyone has some fine-tunes.
Seems to be an issue with the repo currently. I'm not able to pull any custom nodes for Stable Audio Sampler, either through ComfyUI or manually cloning it.
Fails to install nodes due to pedalboard version 0.7.4 being unavailable? It looks like 0.8.2 is the next version available. Can't seem to figure out where to edit requirements for stable-audio-tools because its in that package .py somewhere else but I can't find it? Any ideas?
My question concerns the foundational knowledge base. Does LK's training already include the mechanics of the various rules systems, such as 5e, Pathfinder, Black Hack, and others mentioned on the website?
This is awesome thanks again. Question, how do you get the new Sampler node? can't seem to have it working on your workflow file, and if I try to repro I can't find the new Sampler node with the extra inputs? maybe I missed something haha.
If you are using Manager you can just find it by search StableAudio and you should see it. Also double click on the window in comfyui to bring up node search. Then you can also find the node itself that way.
Too cool! Just few days ago I was like, "I wish there were a comfy node for Stable Audio" and yay, here you are! Thank you so much for doing this! Is there a way to prompt with audio?
Very cool this is breaking though now. Will definitely be checking back on this in 6 months and see where we've got too. Really need clear settings for bpm & scale before I can motivate myself to start dabbling.
The results seem pretty terrible, especially vs closed source options, but I'm glad this has finally been released. It should improve once in the community's hands, like the generic SD models. Those are usually the worst models but people fine tune them and *those* models are a lot better.