LukeG89
LukeG89
There are many people that are having this issue with Florence2 of ComfyUI_LayerStyle: https://github.com/chflame163/ComfyUI_LayerStyle/issues?q=is%3Aissue%20state%3Aopen%20TypeError%3A%20%27NoneType%27%20object%20is%20not%20callable You should try using [ComfyUI-Florence2](https://github.com/kijai/ComfyUI-Florence2) instead.
Or you can try downgrading `transformers`: https://github.com/chflame163/ComfyUI_LayerStyle/issues/322#issuecomment-2413785020 (the version in the comment is wrong apparently) Open terminal inside **python_embeded** folder and run this command to downgrade transformers to 4.43.2 (the...
> that gives me this error ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following...
@RogerWeihrauch The Checkpoint you used has no baked VAE in it. You need to load a VAE and connect it to VAEDecode. 
You probably used a model (maybe a ControlNet) that is not compatible with the Checkpoint. You need to check if all models are of the same type (for example all...
> How can I be/make sure that a ckpt HAS integrated a VAE? @RogerWeihrauch I honestly don't know if there is a way to know. Regarding your case, I just...
Could it be that GGUF model is not compatible with Controlnet Union? Have you tried with the Unet model instead?
I think it goes out of memory because you set 19 in "repeat", so a batch of 19 images (and with Flux is a lot): 
You can go to [CivitAI](https://civitai.com/), there are plenty of models to choose from. I would also suggest this YT channel: [Latent Vision](https://www.youtube.com/@latentvision) He is the developer of [IPAdapter Plus](https://github.com/cubiq/ComfyUI_IPAdapter_plus) node...
> Update, seems like after [#10899](https://github.com/comfyanonymous/ComfyUI/pull/10899) something has been fixed but now I get another error: > > `Expected size for first two dimensions of batch2 tensor to be: [64,...