Flux.1-Dev-Stamp-Art-LoRA

Prompt

stam9, Captured on a light green background is a postage stamp that reads "19" and "Starwberry". The stamp features two red strawberries with green stems and green leaves. The strawberries have small white dots on them. Below the strawberries, the word "starwberry" is written in black letters. The stamp has a scalloped border around the edges.

Prompt

stam9, Captured from a high-angle perspective on a vibrant red backdrop, a square-shaped postage stamp features a red tomato with a green stem and a green leaf on top of it. The tomato is drawn in a cartoon-like fashion, with the word "TOMATO" written below it in a bold, blue font. The background is a light beige, with a rough texture, adding depth to the composition.

Prompt

stam9, Captured from a close-up perspective on a deep royal blue backdrop, a square-shaped postage stamp features a bright yellow banana, slightly peeled, with its smooth skin curling down. The banana is illustrated in a playful cartoon style, with the word "BANANA" written above it in a cheerful, curved red font. The background has a soft gradient of lighter blue shades, creating a subtle sense of movement and contrast.

The model is still in the training phase. This is not the final version and may contain artifacts and perform poorly in some cases.

Model description

prithivMLmods/Flux.1-Dev-Stamp-Art-LoRA

Image Processing Parameters

Parameter	Value	Parameter	Value
LR Scheduler	constant	Noise Offset	0.03
Optimizer	AdamW	Multires Noise Discount	0.1
Network Dim	64	Multires Noise Iterations	10
Network Alpha	32	Repeat & Steps	19 & 1850
Epoch	10	Save Every N Epochs	1

Labeling: florence2-en(natural language & English)

Total Images Used for Training : 24 

Best Dimensions

• 1024 x 1024 (Default)

Setting Up

import torch
from pipelines import DiffusionPipeline

base_model = "black-forest-labs/FLUX.1-dev"
pipe = DiffusionPipeline.from_pretrained(base_model, torch_dtype=torch.bfloat16)

lora_repo = "prithivMLmods/Flux.1-Dev-Stamp-Art-LoRA"
trigger_word = "stam9"  
pipe.load_lora_weights(lora_repo)

device = torch.device("cuda")
pipe.to(device)

Data source

- https://playground.com/

Full Stamp Generation without Backdrop

Stamp ID	Description
stam9	A vibrant orange-yellow smiley face prominently displayed against a stark red backdrop. The face features three black circles, each with a small black dot in the center. The mouth is open, revealing a large white knife sticking out of it. Below the knife, the word "DUCKZILLA" is written in bold, black letters. To the right, black hand-written text reads "It's Chork!" in bold, red font. The background is a combination of white and black, providing a striking contrast to the orange face.
stam9	A medium-sized painting of a margarita glass filled with red liquid, set against a dark blue background. The glass is surrounded by green leaves, with three strawberries on both the left and right sides of the glass. A white flower is centered in the image, with the letter "D" positioned in the upper left corner of the frame. The background features a deep maroon color with a white border.

Trigger words

You should use stam9 to trigger the image generation.

Download model

Weights for this model are available in Safetensors format.

Download them in the Files & versions tab.