CyberRealisticFlux-DF11

Note: Despite the "FP16" in the filename, the original weights in this repo are actually in BF16 instead, which makes them safe for DF11 compression.

For more information (including how to compress models yourself), check out https://huggingface.co/DFloat11 and https://github.com/LeanModels/DFloat11

Feel free to request for other models for compression as well, although models whose architecture I am unfamiliar with might be slightly tricky for me.

How to Use

diffusers

1. Install the DFloat11 pip package (installs the CUDA kernel automatically; requires a CUDA-compatible GPU and PyTorch installed):

   pip install dfloat11[cuda12]
   # or if you have CUDA version 11:
   # pip install dfloat11[cuda11]
   

2. Download the CyberRealistic_Flux_V2.5_FP16-DF11.safetensors file and place it in a local directory of your choice.

3. To use the DFloat11 model, run the following example code in Python:

   import torch
   from diffusers import FluxPipeline, FluxTransformer2DModel
   from dfloat11 import DFloat11Model
   pattern_dict = {
       "transformer_blocks\.\d+" : (
           "norm1.linear",
           "norm1_context.linear",
           "attn.to_q",
           "attn.to_k",
           "attn.to_v",
           "attn.add_k_proj",
           "attn.add_v_proj",
           "attn.add_q_proj",
           "attn.to_out.0",
           "attn.to_add_out",
           "ff.net.0.proj",
           "ff.net.2",
           "ff_context.net.0.proj",
           "ff_context.net.2",
       ),
       "single_transformer_blocks\.\d+" : (
           "norm.linear",
           "proj_mlp",
           "proj_out",
           "attn.to_q",
           "attn.to_k",
           "attn.to_v",
       )
   }
   with no_init_weights():
     transformer = FluxTransformer2DModel.from_config(
         FluxTransformer2DModel.load_config(
             "black-forest-labs/FLUX.1-dev",
             subfolder="transformer"
         ),
         torch_dtype=torch.bfloat16
     ).to(torch.bfloat16)
   
   pipe = FluxPipeline.from_pretrained(
       "black-forest-labs/FLUX.1-dev",
       transformer=transformer,
       torch_dtype=torch.bfloat16
   )
   DFloat11Model.from_single_file('CyberRealisticFlux-DF11/CyberRealistic_Flux_V2.5_FP16-DF11.safetensors', device='cpu', bfloat16_model=pipe.transformer, pattern_dict=pattern_dict) # Make sure to download the file first, and edit the filepath accordingly
   pipe.enable_model_cpu_offload()
   prompt = "A beautiful woman with fair skin and long, messy blonde hair styled in a high ponytail with dramatic, face-framing bangs, her green eyes glinting under a cheeky, smirking expression, subtle head tilt adds playful confidence, she wears a translucent, form-fitting dress that clings tastefully to her silhouette, heavy yet refined makeup accentuating her eyes and lips, captured in a blend of long shot and medium close-up for cinematic focus, bathed in warm sunset glow casting soft golden highlights on her skin and hair, rich depth of field, high-detail realism with sensual elegance and atmospheric lighting"
   image = pipe(
       prompt,
       guidance_scale=3.5,
       num_inference_steps=30,
       max_sequence_length=256,
       generator=torch.Generator("cpu").manual_seed(0)
   ).images[0]
   image.save("CyberRealistic_Flux_V2.5_FP16-DF11.png")
   

ComfyUI

Refer to this model instead.