LogoIconEmojiMOE-v0.2-FLUX.1-dev-LoRA

Prompt

LogoIconEmojiMOE, surprised P3nGu1nZz, a microsoft 3D Fluent Emoji of a blue surprised P3nGu1nZz cat with a surprised expression on its face, set against a black background. It appears to be an emoji of a cat, with its eyes wide open and its mouth slightly agape. Its fur is a bright blue color, and its ears are perked up.

Prompt

LogoIconEmojiMOE, a microsoft 3D Fluent Emoji of a green goblin, smiley face with a big smile on its goblin face, inside his mouth is colored magenta, 3D rendered goblin emoji is against a light pastel orange background.

Prompt

LogoIconEmojiMOE, a Microsoft 3D Fluent Emoji of a retro pixel-art panda, composed of tiny cubic voxel textures. Its eyes are charmingly oversized, with a warm, nostalgic glow like an old arcade screen. The pixelated fur has subtle shading, giving it a soft 3D feel. This adorable voxel-style panda sits happily against a gentle pastel lavender background.

Prompt

LogoIconEmojiMOE a baby blue ui/ux button, that is rectangular with rounded corners, with black background with the words "Support Me On Ko-fi" written in black THICK FAT RETRO font,"Ko-fi" is written after the mug icon and ko-fi is full bold text compared to the "Support me on" text, accompanied by a coffee mug with a heart symbol next to Ko-fi all written on the button which encapsulates the entire canvas.

Prompt

LogoIconEmojiMOE, a 3D isometric view, rendered cluster of x4 identical, upright Nvidia Geforce GTX TITAN XP graphics cards plugged into a motherboard, with green lights illuminating them, set against a black background.

Prompt

LogoIconEmojiMOE, 3D icon collage, Icon banner, with the title "What will you create?" on the screen, along with various 3D rendered icons and text with modern gradients, transparency, and ray casting. It appears to be a screenshot from a stream deck website banner, suggesting that the user is in the process of creating something, and that the banner is providing relevant stream deck icons.

Prompt

LogoIconEmojiMOE, a modern, swirly cursive font logo, the text says "ModdedAgents" where Modded is cream colored, and Agents is orange colored, the ModdedAgents logo is a professional text logo for a modern ai company.

Prompt

LogoIconEmojiMOE, modern app icon featuring the Adobe Photoshop Ps logo enclosed in a blue circle with a long shadow, set against a dark background for a sleek, polished appearance. The design incorporates four unique variations: Kaleidoscope, Pulse, Swirl, Twist, and Warp, each dynamically reinterpreting the logo while maintaining its recognizable branding. Kaleidoscope fractures the Ps logo into symmetrical, mirrored segments, creating a mesmerizing pattern. Pulse amplifies energy waves radiating from the logo for a rhythmic, bold visual.Swirl distorts the circular frame with fluid, spiraling motion. Twist bends and stretches the form into a dynamic, sleek reconfiguration. Warp shifts the structure with futuristic distortions, giving it a cutting-edge effect. Each variation enhances the modern logo aesthetic, blending the clean, professional Photoshop branding with abstract artistic transformations to create a visually striking and cohesive design pack.

Prompt

LogoIconEmojiMOE circular, modern, app icon, the Adobe Acrobat Reader logo in the center of a maroon circle with a long shadow, set against a dark background.

Utilizing a mixture of experts dataset, of 100 unique Logo's, Icon's, and Emoji's including the microsoft fluentui emojis, some iconic software company logos, as well os some synthetic img examples, which were generated by other models such as the Borcherding/FLUX.1-dev-LoRA-HuggingFaceIcons-v0.1 LogoIconEmojiMOE-v0.2-FLUX.1-dev-LoRA was finetuned on an NVIDIA Titan XP with 12GB of VRAM using Fluxgym

Trigger words

You should use LogoIconEmojiMOE to trigger the image generation.

Installation

First, ensure you have FLUX.1 [dev] installed:

# Install
pip install -U diffusers

# Loading with Diffusers
import torch
from diffusers import FluxPipeline

# Load base model
pipe = FluxPipeline.from_pretrained("black-forest-labs/FLUX.1-dev", torch_dtype=torch.bfloat16)
pipe.enable_model_cpu_offload()

# Load and apply LoRA weights
pipe.load_lora_weights("Borcherding/LogoIconEmojiMOE-v0.2-FLUX.1-dev-LoRA")

# Generate image
prompt = "HuggingFaceIcons a yellow smiley face sun with a book in its hands, surrounded by planets and stars in a dark background."
image = pipe(
    prompt,
    height=1024,
    width=1024,
    guidance_scale=3.5,
    num_inference_steps=50,
    max_sequence_length=512,
    generator=torch.Generator("cpu").manual_seed(0)
).images[0]
image.save("autumn-scene.png")

# Unload LoRA weights if needed
pipe.unload_lora_weights()

# Optional: Merge LoRA weights permanently
pipe = FluxPipeline.from_pretrained("black-forest-labs/FLUX.1-dev", torch_dtype=torch.bfloat16)
pipe.load_lora_weights("Borcherding/LogoIconEmojiMOE-v0.2-FLUX.1-dev-LoRA")
pipe.merge_lora_weights()
pipe.save_pretrained("LogoIconEmojiMOE-v0.2-FLUX.1-dev-LoRA")

Trigger Word

Use trigger word "LogoIconEmojiMOE" in your prompts to activate the LoRA. Best results achieved when placing the trigger word at the start of the prompt. For optimal prompting, refer directly to the image annotations in the training data. The style variations, similar to the trigger word, will trigger the model to generate in that style type, and this is what makes LogoIconEmojiMOE-v0.2-FLUX.1-dev-LoRa a Mixture of experts (MOE). Eventually I may try merging this models with others to see the effects.

Here is the training data which provides a direct insight into the prompting techniques you will want to apply:

LogoIconEmojiMOE

Limitations

This is a LoRA for FLUX.1 [dev] and requires the base model to function Finetuned on a small dataset (100 images). Inherits base limitations from FLUX.1 [dev]

License This LoRA falls under the same licensing terms as FLUX.1 [dev]. Please refer to the base model's license for usage terms. Credits

Base model: FLUX.1 [dev] by Black Forest Labs LoRA training and development: [Borcherding at BorchInk]

Download model

Weights for this model are available in Safetensors format.

Download them in the Files & versions tab.