ERROR lora diffusion_model.img_in.weight shape '[3072, 384]' is invalid for input of size 196608 I always get errors in the console with Turbo Lora and FILL models, I think the lora is not working, the quality is quite low. do you not have this error? do you know how to fix it?
Not sure if you can maybe help with an issue I’m running across , first let me say the workflow is amazing having alot of fun with it , I run a higher end pc 3090rtx 64gb ram , when I try to work on a photo at higher res I can “warning ran out of memory” I’ve searched around and can’t find a fix to this
Yes you can use non GGUF version just delete the GGuF checkpoint node and replace it with a normal flux checkpoint node and connect the node appropriately.
Yes, exactly no need for prompt in this workflow, actually is better to allow the Redux work like prompt its way better than using text prompt as it uses the Image as its prompt.
ERROR lora diffusion_model.img_in.weight shape '[3072, 384]' is invalid for input of size 196608
I always get errors in the console with Turbo Lora and FILL models, I think the lora is not working, the quality is quite low. do you not have this error? do you know how to fix it?
You're a class act. Great tutorials. Love your "flow" and great voice :familiar, no attitude and concise! Def one on my favorite go to's! Thank you!
Wow, thank you! I appreciate it! Happy new year!
Not sure if you can maybe help with an issue I’m running across , first let me say the workflow is amazing having alot of fun with it , I run a higher end pc 3090rtx 64gb ram , when I try to work on a photo at higher res I can “warning ran out of memory” I’ve searched around and can’t find a fix to this
hey, can i use non gguf version? how to incorporate non gguf version in this workflow
Yes you can use non GGUF version just delete the GGuF checkpoint node and replace it with a normal flux checkpoint node and connect the node appropriately.
most of the results look pretty fake like the hair was copy pasted on. any way to make the results more realistic looking?
Nice workflow! Thank you for sharing. Looks like there is no need to type my prompt. Why? Wouldn't that produce better results?
Yes, exactly no need for prompt in this workflow, actually is better to allow the Redux work like prompt its way better than using text prompt as it uses the Image as its prompt.
Please share work flow, great video 🎉
Here is the link to the workflow: civitai.com/models/1088986/ultimate-flux-fill-inpainting-redux-model-ft-manual-masking-workflow
@@AiMotionStudio Thank you very much.
Great! Perfect! Sperb!👍👍👍👍👍👍👍👍👍👍👍👍👍👍👍👍👍👍👍👍👍👍
🎉🎉🎉🎉🎉
th-cam.com/video/V8bAIKA552U/w-d-xo.html