Thank You, Thank you, Thank you for fixing this workflow. I was trying to make it work, but it had trouble recognizing the Flux1_Dev_fp8. I'm sure it works for others, but it didn't work for me. Keep in mind I can use my Flux1_Dev_fp8 on other workflows. FINALLY I can flux outpaint. I tried four other inferior workflows, but I knew I really wanted this one to work. And it now does, Thank you!😁
Hi there, i have and isuue with this workflow and i would like to know if u can help me. The thing is that im getting blurry images coming out from the restore detail, do you know why can be this happen? thanks...fantastic workflow.
Thanks for the Video. I'm getting this error: LayerUtility: ImageBlendAdvance V2 `np.asfarray` was removed in the NumPy 2.0 release. Use `np.asarray` with a proper dtype instead.
Can you make a tutorial on flux inpainting with absolutely no visible seams? for example to inpaint complex concepts that only flux can generate (and SDXL can't)
We know you got this workflow we all nodes pre-placed, but we are facing a problem. I loaded the workflow, but most nodes come out red with nothing on them. How do we fix this? Would you tell us how to make the nodes show their content. Without knowing that, the workflow is useless to newbies like me.
If you have red nodes, just go to comfy manager and there is an option of 'Install missing custom nodes', it will automatically install all missing nodes
Thanx. But i'm too lazy. I connect Florence with Datail caption to original prompt input and another one (but with More detailed caption preset) to Flux prompt input;)
RTX 4060 Laptop GPU. One thing mentioned in the video is slightly wrong, the normal out painting takes around 2-2.5 minutes for full process, but you can replace the juggernautXL with Dreamshaper Turbo with 8 steps and the time can be reduced more.
Fantastic work....in my case it would be great to have this workflow in a Flow style. Did you ever think about it? thanks
Thank You, Thank you, Thank you for fixing this workflow. I was trying to make it work, but it had trouble recognizing the Flux1_Dev_fp8. I'm sure it works for others, but it didn't work for me. Keep in mind I can use my Flux1_Dev_fp8 on other workflows. FINALLY I can flux outpaint. I tried four other inferior workflows, but I knew I really wanted this one to work. And it now does, Thank you!😁
Hi there, i have and isuue with this workflow and i would like to know if u can help me. The thing is that im getting blurry images coming out from the restore detail, do you know why can be this happen? thanks...fantastic workflow.
Thanks for the Video. I'm getting this error:
LayerUtility: ImageBlendAdvance V2
`np.asfarray` was removed in the NumPy 2.0 release. Use `np.asarray` with a proper dtype instead.
Can you share some in painting workflow that blends smoothly? My inpainting looks like it's photoshopped and not natural.
Can you make a tutorial on flux inpainting with absolutely no visible seams? for example to inpaint complex concepts that only flux can generate (and SDXL can't)
We know you got this workflow we all nodes pre-placed, but we are facing a problem. I loaded the workflow, but most nodes come out red with nothing on them. How do we fix this? Would you tell us how to make the nodes show their content. Without knowing that, the workflow is useless to newbies like me.
If you have red nodes, just go to comfy manager and there is an option of 'Install missing custom nodes', it will automatically install all missing nodes
Friend, is this also possible to do in Forge WebUI?
I need help with my workflow. how can i connect?
Hi, what problem are you facing ?
Thanx. But i'm too lazy. I connect Florence with Datail caption to original prompt input and another one (but with More detailed caption preset) to Flux prompt input;)
How is that? can u share your workflow please? thanks
Which version of graphic card you are using?
RTX 4060 Laptop GPU. One thing mentioned in the video is slightly wrong, the normal out painting takes around 2-2.5 minutes for full process, but you can replace the juggernautXL with Dreamshaper Turbo with 8 steps and the time can be reduced more.