lllyasviel
Same lora loading multiple times
The same lora is being loaded twice as you can see:
I have called the loras just once in the prompt. I am using PonyDiffusionV6 with its corresponding VAE. The loras are compatible with this model last time i checked. Also i am using Clip skip = 2.
How much VRAM do you have, and what's the GPU weights setting at the top of the screen set to?
I have 16 gb ram, using macbook pro m1 pro. But i had no problem running other models and loras. I even ran Flux on this device. GPU Weights=15360MB
Yeah you're fine hardware-wise. I didn't notice it before, but It's loading the UNET then the CLIP weights for each Lora. I tried reproducing this, but my setup only loads the UNET weights. Are you loading any custom CLIP checkpoints or anything like that? Perhaps an extension somewhere doing something?
No, i am not using any custom CLIP checkpts or extensions. I just downloaded the lora and then added it to the prompt just how its supposed to be added and then this stuff happened
Looks like its specific to the Loras you're using. I grabbed one of the loras from your screenshot to check and saw the same thing:
[LORA] Loaded D:\AI\stable-diffusion-webui-forge\models\Lora\pony\Western_art_style_-_Melkor_Mancin__Riz_draws_-_combined_style-000037.safetensors for KModel-UNet with 722 keys at weight 1.0 (skipped 0 keys) with on_the_fly = False
[LORA] Loaded D:\AI\stable-diffusion-webui-forge\models\Lora\pony\Western_art_style_-_Melkor_Mancin__Riz_draws_-_combined_style-000037.safetensors for KModel-CLIP with 264 keys at weight 1.0 (skipped 0 keys) with on_the_fly = False
My semi-educated guess is that they trained the text encoder (CLIP) when making the Lora. Someone more knowledgeable than I can confirm if that's the case. At any rate, its working as expected and is nothing to worry about.
So then what should i do? When I try the 3 Loras combined I get an output, but the overall design is not as I intended. I downloaded the loras from civitai and also tried them there and they worked fine, but locally they are giving me problems