B R I A 2.3

briaai

Introduction

BRIA 2.3 is a text-to-image model developed by BRIA AI, designed for commercial applications with a focus on legal compliance and ethical content generation. It offers full legal liability coverage and is trained on a licensed dataset, ensuring high-quality, compliant imagery for business purposes.

Architecture

BRIA 2.3 is a latent diffusion text-to-image model. It utilizes a patented attribution engine for compensating data partners and is enterprise-ready with customizable technology options, allowing access to source code and weights for tailored business solutions.

Training

The model is trained exclusively on a professional-grade, licensed dataset, which excludes copyrighted materials, ensuring ethical and legally compliant outcomes.

Guide: Running Locally

  1. Install Required Packages:

    pip install diffusers
    
  2. Load the Model:

    from diffusers import DiffusionPipeline
    import torch
    
    pipe = DiffusionPipeline.from_pretrained("briaai/BRIA-2.3", torch_dtype=torch.float16, use_safetensors=True)
    pipe.force_zeros_for_empty_prompt = False
    pipe.to("cuda")
    
  3. Generate an Image:

    prompt = "A portrait of a Beautiful and playful ethereal singer, golden designs, highly detailed, blurry background"
    negative_prompt = "Logo,Watermark,Text,Ugly,Morbid,Extra fingers,Poorly drawn hands,Mutation,Blurry,Extra limbs,Gross proportions,Missing arms,Mutated hands,Long neck,Duplicate,Mutilated,Mutilated hands,Poorly drawn face,Deformed,Bad anatomy,Cloned face,Malformed limbs,Missing legs,Too many fingers"
    
    images = pipe(prompt=prompt, negative_prompt=negative_prompt, height=1024, width=1024).images[0]
    

Tips for Inference:

  • Set pipe.force_zeros_for_empty_prompt = False.
  • Use negative prompts for better results.
  • Use 30-50 steps for BRIA 2.3; use 8 steps for the Fast model.
  • Use a guidance_scale of 5.0 or 7.5 for BRIA 2.3; 1.0 for the Fast model.

Cloud GPU recommendation: Consider using cloud GPU services like AWS, Google Cloud, or Azure for optimal performance.

License

BRIA 2.3 operates under commercial licensing terms, requiring a purchase for access and use. The licensing covers legal liabilities and ensures royalty management. For more details, refer to BRIA AI's terms and conditions.

More Related APIs in Text To Image