--- tags: - text-to-image - lora - diffusers - template:diffusion-lora widget: - text: 'EBook Cover, An eye-level view of an ancient ruins, with the words "ECHOES OF THE PAST" written in gold in the center of the image. The ruins are made of stone, and the pillars are adorned with intricate carvings. In the background, a mountain can be seen. The sky is a muted yellow, and at the top, the word "By Cesar Blair" is written in a bold black font.' output: url: images/E1.png - text: 'EBook Cover, Captured in a close-up, eye-level perspective, the cover of the book "Devond The SKIES" features a torn brown background with a white outline in the center of the cover. The title, "devond the skies" is written in a bold, white font. Below the title, the author name, "By Steven Brown," is in a smaller font beneath the title.' output: url: images/E2.png - text: 'EBook Cover, A poster of a woman in a black dress, standing in front of a circular window. The womans face is tilted upward, and her eyes are closed. The background of the poster is black, and the womans hair is wet. The window is cracked, and there is a light shining through the window. Below the window, the words "The INVADERS" are written in blue letters, and "By James Stall" is written in white letters.' output: url: images/E3.png base_model: black-forest-labs/FLUX.1-dev instance_prompt: EBook Cover license: creativeml-openrail-m --- # EBook-Creative-Cover-Flux-LoRA **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/EBook-Creative-Cover-Flux-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 | 23 & 2240| | Epoch | 15 | Save Every N Epochs | 1 | Labeling: florence2-en(natural language & English) Total Images Used for Training : 17 [ Hi-RES ] ## Best Dimensions - 1024 x 1024 (Default) - 768 x 1024 ## 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/EBook-Creative-Cover-Flux-LoRA" trigger_word = "EBook Cover" pipe.load_lora_weights(lora_repo) device = torch.device("cuda") pipe.to(device) ``` ## Data source - https://playground.com/ ## Sample 768 x 1024 ![prithivMLmods/EBook-Creative-Cover-Flux-LoRA](images/E4.png) | **Prompt** | |-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| | Close-up, eye-level view of a mysterious eBook cover titled "The Lost Path." The background is a worn, dark blue, with faint compass markings across the cover. The title, "The Lost Path," is written in a bold, metallic silver font, slightly distressed. Beneath the title, the author's name, "By Caroline Grey," appears in a smaller, elegant font in white. | ## Trigger words You should use `EBook Cover` to trigger the image generation. ## Download model Weights for this model are available in Safetensors format. [Download](/prithivMLmods/EBook-Creative-Cover-Flux-LoRA/tree/main) them in the Files & versions tab.