Flux.1-Dev-Movie-Boards-LoRA

Prompt

movieboard, A black and white photograph of a man in a space suit is seen on a page of a magazine. The man is wearing a white helmet on his head, and is standing in a field of tall grass. The background of the photograph is a bright yellow circle, and the word "InterSTAR" is written in bold black letters at the top of the page. The number "20" is in the upper right corner of the frame.

Prompt

movieboard, a vibrant orange-brown poster is adorned with a black frame. The poster features a black figure of a robot, with the words "Iron Man" written in bold black letters at the top of the poster. The robot is facing to the right, with its arms outstretched. The backdrop is a stark white wall, with a textured black border. The bottom left corner of the frame has the number "08" in white letters.

Prompt

movieboard, Captured from a high-angle perspective on a white backdrop, a black framed poster is adorned with the words "SPIDER-MAN" in bold, black font. The poster features a black silhouette of a spider-man in a green jacket, blue pants, and red and white shoes. The man is jumping in the air, with his arms out to the side. His legs are bent at the knees, and his feet are crossed at the bottom of the frame. The backdrop is a light gray, with a white outline of a cityscape in the background. The text "into the Spider-Verse" is written in bold black font at the top of the poster. Below the poster, the date "1/2018" is printed in white font.

Prompt

movieboard, Captured on a white poster, a man in a gray jacket and black pants is holding a gray bag in his left hand. He is standing in front of a large red wave. Above the man, the word "Joker" is written in bold black letters. To the top left of the poster, there is a rainbow colored logo.

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-Movie-Boards-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	16 & 2100
Epoch	14	Save Every N Epochs	1

Labeling: florence2-en(natural language & English)

Total Images Used for Training : 11

Best Dimensions

768 x 1024 (Best)
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-Movie-Boards-LoRA"
trigger_word = "movieboard"  
pipe.load_lora_weights(lora_repo)

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