FluxBooru-v0.2-LoKr-1.6B

flux-booru-v0.2

This is a LyCORIS adapter derived from terminusresearch/flux-booru-v0.2.

The main validation prompt used during training was:

A photo-realistic image of a cat

Validation settings

• CFG: 4.0
• CFG Rescale: 0.0
• Steps: 20
• Sampler: None
• Seed: 42
• Resolution: 1024x1024

Note: The validation settings are not necessarily the same as the training settings.

You can find some example images in the following gallery:

<Gallery />

The text encoder was not trained. You may reuse the base model text encoder for inference.

Training settings

• Training epochs: 0
• Training steps: 8000
• Learning rate: 0.0001
• Max grad norm: 1.0
• Effective batch size: 32
  • Micro-batch size: 4
  • Gradient accumulation steps: 1
  • Number of GPUs: 8
• Prediction type: flow-matchingNone
• Rescaled betas zero SNR: False
• Optimizer: adamw_bf16
• Precision: Pure BF16
• Quantised: No
• Xformers: Not used
• LyCORIS Config:

{
    "algo": "lokr",
    "multiplier": 1.0,
    "linear_dim": 10000,
    "linear_alpha": 1,
    "factor": 4,
    "apply_preset": {
        "target_module": [
            "FluxTransformer2DModel"
        ],
        "module_algo_map": {
            "Attention": {
                "factor": 4
            },
            "FeedForward": {
                "factor": 2
            }
        }
    }
}

Datasets

celebrities

• Repeats: 0
• Total number of images: ~3960
• Total number of aspect buckets: 22
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

comicstrips

• Repeats: 0
• Total number of images: ~1624
• Total number of aspect buckets: 48
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

movieposters

• Repeats: 0
• Total number of images: ~3144
• Total number of aspect buckets: 19
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

normalnudes

• Repeats: 0
• Total number of images: ~1304
• Total number of aspect buckets: 22
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

propagandaposters

• Repeats: 0
• Total number of images: ~1456
• Total number of aspect buckets: 19
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

guys

• Repeats: 100
• Total number of images: ~464
• Total number of aspect buckets: 19
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

pixel-art

• Repeats: 100
• Total number of images: ~1224
• Total number of aspect buckets: 18
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

signs

• Repeats: 0
• Total number of images: ~448
• Total number of aspect buckets: 14
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

moviecollection

• Repeats: 0
• Total number of images: ~2016
• Total number of aspect buckets: 24
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

bookcovers

• Repeats: 0
• Total number of images: ~936
• Total number of aspect buckets: 15
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

nijijourney

• Repeats: 0
• Total number of images: ~1520
• Total number of aspect buckets: 27
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

experimental

• Repeats: 0
• Total number of images: ~3152
• Total number of aspect buckets: 24
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

ethnic

• Repeats: 0
• Total number of images: ~3184
• Total number of aspect buckets: 22
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

sports

• Repeats: 0
• Total number of images: ~880
• Total number of aspect buckets: 17
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

gay

• Repeats: 0
• Total number of images: ~1176
• Total number of aspect buckets: 21
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

architecture

• Repeats: 0
• Total number of images: ~4472
• Total number of aspect buckets: 25
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

shutterstock

• Repeats: 0
• Total number of images: ~21200
• Total number of aspect buckets: 17
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

cinemamix-1mp

• Repeats: 0
• Total number of images: ~7440
• Total number of aspect buckets: 5
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

nsfw-1024

• Repeats: 100
• Total number of images: ~10888
• Total number of aspect buckets: 1
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

anatomy

• Repeats: 5
• Total number of images: ~16520
• Total number of aspect buckets: 2
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

bg20k-1024

• Repeats: 0
• Total number of images: ~89480
• Total number of aspect buckets: 8
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

yoga

• Repeats: 0
• Total number of images: ~3696
• Total number of aspect buckets: 20
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

photo-aesthetics

• Repeats: 0
• Total number of images: ~33288
• Total number of aspect buckets: 8
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

text-1mp

• Repeats: 5
• Total number of images: ~13312
• Total number of aspect buckets: 2
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

sfwbooru

• Repeats: 0
• Total number of images: ~639104
• Total number of aspect buckets: 21
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None

Inference

import torch
from diffusers import DiffusionPipeline
from lycoris import create_lycoris_from_weights

model_id = 'terminusresearch/flux-booru-v0.2'
adapter_id = 'pytorch_lora_weights.safetensors' # you will have to download this manually
lora_scale = 1.0
wrapper, _ = create_lycoris_from_weights(lora_scale, adapter_id, pipeline.transformer)
wrapper.merge_to()

prompt = "A photo-realistic image of a cat"

pipeline.to('cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu')
image = pipeline(
    prompt=prompt,
    num_inference_steps=20,
    generator=torch.Generator(device='cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu').manual_seed(1641421826),
    width=1024,
    height=1024,
    guidance_scale=4.0,
).images[0]
image.save("output.png", format="PNG")