Any SD 1.5 equivalent?

haha... now I need to re-generate all those pics I used 1.3 (or 1.2 only) on haha... this has become like the standard resource I am loading for just about anything... major thanks and kudos!

Please make an Midjourney v5.2 sdxl model instead of lora

is there a reason you deleted v1.3?

I was wrong 2.0 still says segmentation fault error:<

v1.2 works fine, but v2.0 with everything else the same returns 'StableDiffusionXLPipeline' object has no attribute 'diffusion' and then fails to load.

Use v2.0 full weight and add v1.2 in 0.1 - 0.8 of weight.

If I try the 2 loras together the app crashes immediately (Draw Things for OS) !!

The v2.0 and the v1.3 don't work in Diffusers 0.24. Error loading.

Using Stable Diffusion, v1.2 and 1.3 work fine but I get this error with v2.0: Error: Could not load the lora model! Reason: 'StableDiffusionXLPipeline' object has no attribute 'diffusion'

I am not able to download this in my kaggle it says 404 bad errro even though I have tried the API it was not downloading

import os

# Define the models to download and their respective destinations

models = {

#"Jellyfishforest.safetensors": {"url": "https://civitai.com/api/download/models/114163", "dest": "temp"},

"MJ52_v2.0.safetensors": {"url": "https://civitai.com/api/download/models/678485?type=Model?token=", "dest": "temp"},

#"The_Serpent_God.safetensors": {"url": "https://civitai.com/models/41945/naga-the-serpent-god", "dest": "temp"},

#"ral-gamingrig-sdxl.safetensors": {"url": "https://civitai.com/api/download/models/614981?type=Model&format=SafeTensor", "dest": "temp"},

#"pokemon_v3_offset.safetensors": {"url": "https://civitai.com/api/download/models/73529?type=Model&format=SafeTensor", "dest": "temp"},

#"TransformersStyleXL-000009.safetensors": {"url": "https://civitai.com/api/download/models/245043?type=Model&format=SafeTensor", "dest": "working"},

#"comic_book_illustration.safetensors": {"url": "https://civitai.com/api/download/models/271577?type=Model&format=SafeTensor", "dest": "working"},

#"dragonborn_offset.safetensors": {"url": "https://civitai.com/api/download/models/123280?type=Model&format=SafeTensor", "dest": "working"},

#"jim_lee_offset_right_filesize.safetensors": {"url": "https://civitai.com/api/download/models/10580?type=Model&format=SafeTensor&size=full&fp=fp16", "dest": "working"},

#"FWXL.safetensors": {"url": "https://civitai.com/api/download/models/660798?type=Model&format=SafeTensor", "dest": "temp"},

"Future_Warfare_SDXL.safetensors": {"url": "https://civitai.com/api/download/models/181070?type=Model&format=SafeTensor", "dest": "temp"},

#Add new models here, e.g.:

# "model_name.safetensors": {"url": "model_url", "dest": "temp"},

# "another_model.safetensors": {"url": "another_model_url", "dest": "working"},

}

# Create directories if they don't exist

!mkdir -p /kaggle/temp/Lora/models

!mkdir -p /kaggle/working/stable-diffusion-webui-forge/models/Lora/temp_Lora_models

# Download and install models

for model_name, info in models.items():

if info["dest"] == "working":

dest_dir = "/kaggle/working/stable-diffusion-webui-forge/models/Lora/temp_Lora_models"

elif info["dest"] == "temp":

dest_dir = "/kaggle/temp/Lora/models"

# Check if model is already downloaded in either temp or working directory

if not (os.path.exists(f"/kaggle/temp/Lora/models/{model_name}") or os.path.exists(f"/kaggle/working/stable-diffusion-webui-forge/models/Lora/temp_Lora_models/{model_name}")):

# Download the model

get_ipython().system(f'wget -O "{dest_dir}/{model_name}" "{info["url"]}"')

print(f"Downloaded {model_name} to {dest_dir}")

else:

print(f"{model_name} already exists in either temp or working directory")

continue # Skip to next model

# Create symlink to temp Lora models directory (if temp)

if info["dest"] == "temp":

symlink_dest = f"/kaggle/working/stable-diffusion-webui-forge/models/Lora/temp_Lora_models/{model_name}"

if os.path.exists(symlink_dest) and not os.path.isdir(symlink_dest):

os.remove(symlink_dest)

try:

!ln -s {dest_dir}/{model_name} {symlink_dest}

print(f"Created symlink for {model_name}")

except Exception as e:

print(f"Failed to create symlink for {model_name}: {str(e)}")

please help me with this one

So basically none of the image below are made with this LORa...

2.0 doesn't change the image

Works perfectly and as intended! Thank you so much for making this. There is however something that pop's up inside my comfyui terminal. It could be the source on why others are having a problem with it. Here is the output.

MJ52_v2.0.safetensors, ComfyUI: v0.3.9-21-g73e0498 (2024-12-24), Power Lora Loader (rgthree):
lora key not loaded: diffusion_model_output_blocks_2_2_conv.alpha

lora key not loaded: diffusion_model_output_blocks_2_2_conv.lora_down.weight

lora key not loaded: diffusion_model_output_blocks_2_2_conv.lora_up.weight

lora key not loaded: diffusion_model_output_blocks_5_2_conv.alpha

lora key not loaded: diffusion_model_output_blocks_5_2_conv.lora_down.weight

lora key not loaded: diffusion_model_output_blocks_5_2_conv.lora_up.weight

Other LoRAs in the same node (including MJ52.safetensors) load and work without such messages in the console.

I'm using V1.2 with V2 together and the results are stunning! It's crazy by how much the images overall improve with your Lora applied.

Hello, I'm new to civitai.
I use colab to launch fooocus and i set my guidance scale at 6. Lora weight at 1. But my ronaldo photo is look like plastic, not real like in midjourney, what can i do? I'm using only version 2 not 1 because its too large file to upload in colab

Get an error with the 2.0 version in InvokeAI:

ValueError: Unrecognized SDXL LoRA key prefix: 'diffusion_model_output_blocks_2_2_conv.alpha'.

1.2 works just fine.

So, it seems that this LORA has become unusable since the big Checkpoint update?

At least I can’t generate any images with it anymore. All generations fail… 😢

Is anyone else having this problem?

Can we use this in draw things iOS app? I’m trying 1.2 but the app keeps crashing? 1.3 and 2.0 works fine