Bird of Paradise LoRA Image Generator

Discover the Bird of Paradise LoRA Image Generator, a powerful tool for creating exquisite and detailed images of birds of paradise.

Platform: Replicate

LoRA Image GenerationBird of Paradise CreatorImage InpaintingStyle Transfer

34 runs

H100

License Check Required

🚀Function Overview

A LoRA-based image generation model that creates detailed images of birds of paradise based on text prompts, supporting image-to-image transformations, inpainting, style customization, and adjustable quality/speed configurations.

Key Features

Customizable aspect ratios and dimensions
Image-to-image and mask-guided inpainting modes
Multiple model variants for quality vs. speed tradeoffs
Adjustable LoRA scale for style intensity
Compatibility with additional external LoRA models
Safety checker and output quality controls

Use Cases

•Generating artistic depictions of birds of paradise
•Editing/transforming existing images via inpainting
•Content creation for digital art and design projects
•Concept visualization with specific style parameters

⚙️Input Parameters

prompt

string

Prompt for generated image. If you include the `trigger_word` used in the training process you are more likely to activate the trained object, style, or concept in the resulting image.

image

string

Input image for image to image or inpainting mode. If provided, aspect_ratio, width, and height inputs are ignored.

mask

string

Image mask for image inpainting mode. If provided, aspect_ratio, width, and height inputs are ignored.

aspect_ratio

string

Aspect ratio for the generated image. If custom is selected, uses height and width below & will run in bf16 mode

height

integer

Height of generated image. Only works if `aspect_ratio` is set to custom. Will be rounded to nearest multiple of 16. Incompatible with fast generation

width

integer

Width of generated image. Only works if `aspect_ratio` is set to custom. Will be rounded to nearest multiple of 16. Incompatible with fast generation

prompt_strength

number

Prompt strength when using img2img. 1.0 corresponds to full destruction of information in image

model

string

Which model to run inference with. The dev model performs best with around 28 inference steps but the schnell model only needs 4 steps.

num_outputs

integer

Number of outputs to generate

num_inference_steps

integer

Number of denoising steps. More steps can give more detailed images, but take longer.

guidance_scale

number

Guidance scale for the diffusion process. Lower values can give more realistic images. Good values to try are 2, 2.5, 3 and 3.5

seed

integer

Random seed. Set for reproducible generation

output_format

string

Format of the output images

output_quality

integer

Quality when saving the output images, from 0 to 100. 100 is best quality, 0 is lowest quality. Not relevant for .png outputs

disable_safety_checker

boolean

Disable safety checker for generated images.

go_fast

boolean

Run faster predictions with model optimized for speed (currently fp8 quantized); disable to run in original bf16

megapixels

string

Approximate number of megapixels for generated image

lora_scale

number

Determines how strongly the main LoRA should be applied. Sane results between 0 and 1 for base inference. For go_fast we apply a 1.5x multiplier to this value; we've generally seen good performance when scaling the base value by that amount. You may still need to experiment to find the best value for your particular lora.

extra_lora

string

Load LoRA weights. Supports Replicate models in the format <owner>/<username> or <owner>/<username>/<version>, HuggingFace URLs in the format huggingface.co/<owner>/<model-name>, CivitAI URLs in the format civitai.com/models/<id>[/<model-name>], or arbitrary .safetensors URLs from the Internet. For example, 'fofr/flux-pixar-cars'

extra_lora_scale

number

Determines how strongly the extra LoRA should be applied. Sane results between 0 and 1 for base inference. For go_fast we apply a 1.5x multiplier to this value; we've generally seen good performance when scaling the base value by that amount. You may still need to experiment to find the best value for your particular lora.

💡Usage Examples

Example 1

Input Parameters

{
  "model": "dev",
  "prompt": "Profile view of a majestic bird of paradise with ultra colorful wings and feathers that trail behind its body with ornate and hyper-detailed intricacy, perched on a branch in a lush rainforest environment",
  "go_fast": false,
  "lora_scale": 1,
  "megapixels": "1",
  "num_outputs": 1,
  "aspect_ratio": "16:9",
  "output_format": "png",
  "guidance_scale": 3,
  "output_quality": 100,
  "prompt_strength": 0.8,
  "extra_lora_scale": 1,
  "num_inference_steps": 28
}

Output Results

https://replicate.delivery/xezq/xcMPxgjxkPKWBhkm9zUYvo1gOyHfNlSQAhrFhYJicSJUdoQKA/out-0.png

Quick Actions

Use NowView Documentation

Technical Specifications

Hardware Type: H100
Run Count: 34
Commercial Use: Unknown/Restricted
Platform: Replicate

Related Keywords

LoRA Image GenerationImage InpaintingStyle CustomizationText-to-ImageArtistic DepictionsDigital Art

Related Models

Flux-Dev-LoRA-Mahi

A model for generating and modifying images using LoRA techniques with customization options.

Flux Prehistoric Art

A model for generating images with a prehistoric art style through text prompts and image manipulation features.

Alalpaci Image Generation

A model for image generation and editing using textual prompts and input images with fine-grained control over diffusion parameters.