tags:
- text-to-image
- stable-diffusion
- lora
- diffusers
- template:sd-lora
widget:
- text: >-
c0m1c style vintage comic cover of a psychic woman sitting at a covered
red table, in a haunting manor scene, the illustration style is macabre,
candles in the bacground over a mantle, the psychic is hovering her hands
over a purple crystal ball. Title test reads "Tales from the Flux" in an
eerie type face. In the lower corner show a price of 15 cents and the date
SEP 2024
output:
url: images/ComfyUI_09868_.png
- text: >-
c0m1c style vintage 1930s style comic strip panel of a rich old man
wearing a top hat and dapper suit deep in thought, thinking with a thought
bubble that reads "No FLUX were given?"
output:
url: images/ComfyUI_temp_ipugi_00131_.png
- text: >-
c0m1c style comic book strip panel of two men having a staring contest in
a bar, pop art style, with the thought bubble sayings "Blink." and "You
first."
output:
url: images/ComfyUI_temp_ipugi_00079_.png
- text: >-
c0m1c style comic book strip panel of a glam 1980s superhero woman, with
the speech bubble saying "My spirit animal is a WiFi symbol!" purple wifi
icon on her chest
output:
url: images/ComfyUI_temp_ipugi_00071_.png
- text: >-
c0m1c style comic book strip 2 panels of a woman laughing hysterically,
1950s living room. In the first panel one speech bubble "And then he said
he uses SD3!" over the head of a woman telling her friends a funny joke
In the second panel there are multiple speech bubbles over each woman's
head with the words "HAHA" in various ways, their heads are tilted back in
heavy laughter.
output:
url: images/ComfyUI_temp_ipugi_00085_.png
- text: >-
c0m1c style comic book strip panel of a woman in her living room glam
style, with the speech bubble saying "There's no place like latent space."
everything is pink including her dress
output:
url: images/ComfyUI_temp_ipugi_00076_.png
- text: >-
c0m1c style comic book strip panel of a superhero, with the speech bubble
saying "I am Captain Flux!" with a big black F emblem on his chest
output:
url: images/ComfyUI_temp_ipugi_00068_.png
- text: >-
c0m1c style comic strip panel of a pop art woman fashion model, side
profile, speech above says "OH FLUX!"
output:
url: images/ComfyUI_temp_ipugi_00040_.png
- text: >-
c0m1c style woman with red hair in a laundry room, vintage illustrated ad
for laundry detergent the brand name is FLUX in orange letters the
pre-heading above the logo reads "Where clean clothes meet clean pixels."
The woman is holding a laundry baskest with folded clothes and smiling in
bliss.
output:
url: images/ComfyUI_09863_.png
base_model: black-forest-labs/FLUX.1-dev
instance_prompt: c0m1c, comic book panel
license: creativeml-openrail-m
Retro Comic Flux
Model description
Retro Comic Flux is a Flux LoRA trained on a dataset consisting entirely of public domain images manually sourced, cropped and enhanced. I used ~50 images 3 repeats 15 epochs and 0.0001 learning rate. These were all captioned using Joy Caption Batch.
This is very experimental and I had a difficult time sourcing high quality images of animals and vehicles in this style, so for now it's primarily good with human figures. It work best when you describe the character and a scene for a backdrop. It struggles with hands and eyes sometimes. If something is really not in the dataset it defaults to an ordinary AI illustration style.
USE DEIS SAMPLER FOR THE BEST RESULTS.
I used very similar settings as I have with my previous LoRAs so I won't go into detail for the sake of my time and brevity, low learning rates with less repeats and good data work best for simple styles.
I was careful in my selection of images and preprocessed each image in Photoshop to bring out vivid colors, reduce the yellowing and boost contrast. This helped me to maintain a consistent coloring and I believe it improved the training quality. When I did this I also removed the speech from speech bubbles from some of the training data for more flexibility and to help Flux understand context. You can explicitly state "empty speech bubble" in your prompts and it tends to work. If you want to dive in deeper try using an image captioner to describe a style you want to emulate, the images have rich captions and the model responds well to natural language.
Definitely expect a V2 of this LoRA soon, I have some ideas on how I can improve consistency but I need to find more varied source images.
Trigger keyword: 'c0m1c' 'comic book panel'
Recommended strengths: 0.7 - 0.9
Trigger words
You should use c0m1c
to trigger the image generation.
You should use comic book panel
to trigger the image generation.
Download model
Weights for this model are available in Safetensors format.
Download them in the Files & versions tab.