Yes, even though adding a LoRa will make your style a little more aggressive, it is still possible. To control that, you could use the LoRa strength. Thank you for your uplifting words.
Awesome, but what model of Flux are you using to load in "load Checkpoint", usually I open it in "load diffusion" or "unet loader", thanks we really appreciate it
You are welcome, and you are correct; the process of using your model also works well. I am using the FP8 model saved in the checkpoint point folder, which works similarly to loading an SDXL or SD 1.5 model. For clarification, I have described the process in my previous video here ua-cam.com/video/TWSFej_S_bY/v-deo.html
in your model i keep getting a error message StyleModelApply Sizes of tensors must match except in dimension 1. Expected size 2048 but got size 4096 for tensor number 1 in the list. how do i fix it
it could be a resolution issue between the model and the input dimensions. Try resizing the input to (1024x1024) or 2048x2048) to match the tensor size. Also running the latest versions of ComfyUI and the StyleModelApply node-updates often resolve compatibility problems.
Wow, thank you so much! You may have finally made me a full Comfy convert, something I have resisted so far.
That is wonderful to hear! I am glad the videos had such a big impact. Welcome to the comfy family.
Finally, a step-by-step lesson showing the whole process of connecting nodes, thank you very much man 🙏
I'm glad you found it helpful! You are very welcome. 🙏
Precise details from the scratch which is actually helpful, Thanks a lot!
You're welcome, and thanks a lot! I appreciate your thoughts on this.
Really well explained ! Thanks !
I appreciate your kind words.
I don't know how you figure this stuff out. Some kind of genius. Anyways great video.
Thank you for the high praise. I’m glad you enjoyed the video.
Thank you!
You are welcome!
Awesome, clear and to the point loving your videos man :)
Does it work to mix this with the depth lora for example?
Keep it up!
Yes, even though adding a LoRa will make your style a little more aggressive, it is still possible. To control that, you could use the LoRa strength. Thank you for your uplifting words.
very useful tutorial. thanks.
You are very welcome, I'm glad it was helpful.
good job as always
I appreciate your feedback. Thank you.
thank you so much.
How can we apply style of any image to real images, without changing its composition and face
You are very welcome. I would recommend using ControlNet with Canny guidance to help lock down the composition and facial features.
Nice well explained! Thank you.
You are welcome. Thank you for sharing your thoughts.
Awesome, but what model of Flux are you using to load in "load Checkpoint", usually I open it in "load diffusion" or "unet loader", thanks we really appreciate it
You are welcome, and you are correct; the process of using your model also works well. I am using the FP8 model saved in the checkpoint point folder, which works similarly to loading an SDXL or SD 1.5 model.
For clarification, I have described the process in my previous video here ua-cam.com/video/TWSFej_S_bY/v-deo.html
in your model i keep getting a error message StyleModelApply
Sizes of tensors must match except in dimension 1. Expected size 2048 but got size 4096 for tensor number 1 in the list. how do i fix it
it could be a resolution issue between the model and the input dimensions. Try resizing the input to (1024x1024) or 2048x2048) to match the tensor size.
Also running the latest versions of ComfyUI and the StyleModelApply node-updates often resolve compatibility problems.
Hi as usually good videos i sent you on other videos i want your contact to speak about business work
Hello there, Kindly see my previous replies with the details.
@@goshniiAI Hi, i dont have any replies from you