gen13-rev03-st-flux-lokr-lion-3e-5-bs2-ga3-v06-12100

This is a LyCORIS adapter derived from black-forest-labs/FLUX.1-dev.

The main validation prompt used during training was:

caitlin and roxy sit together on a rooftop under the setting sun, the golden light bathing their faces in a soft glow, caitlin tall, athletic frame is relaxed as she leans back, dressed in a casual tank top that highlights her muscular build, while roxy, much smaller in stature, is perched next to her in a loose-fitting hoodie, her pink hair tousled by the wind, they share a quiet laugh, their bond of friendship bringing out a sense of relaxed femininity in both of them, contrasting their tough exteriors

Validation settings

• CFG: 3.5
• CFG Rescale: 0.0
• Steps: 20
• Sampler: None
• Seed: 42
• Resolution: 768x1344

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

You can find some example images in the following gallery:

Prompt

unconditional (blank prompt)

Negative Prompt

blurry, cropped, ugly

Prompt

caitlin and roxy sit together on a rooftop under the setting sun, the golden light bathing their faces in a soft glow, caitlin tall, athletic frame is relaxed as she leans back, dressed in a casual tank top that highlights her muscular build, while roxy, much smaller in stature, is perched next to her in a loose-fitting hoodie, her pink hair tousled by the wind, they share a quiet laugh, their bond of friendship bringing out a sense of relaxed femininity in both of them, contrasting their tough exteriors

Negative Prompt

blurry, cropped, ugly

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

Training settings

• Training epochs: 55
• Training steps: 12100
• Learning rate: 3e-05
• Effective batch size: 6
  • Micro-batch size: 2
  • Gradient accumulation steps: 3
  • Number of GPUs: 1
• Prediction type: flow-matching
• Rescaled betas zero SNR: False
• Optimizer: optimi-lionweight_decay=1e-3
• Precision: Pure BF16
• Quantised: Yes: int8-quanto
• Xformers: Not used
• LyCORIS Config:

{
    "algo": "lokr",
    "multiplier": 1.0,
    "linear_dim": 10000,
    "linear_alpha": 1,
    "factor": 12,
    "apply_preset": {
        "target_module": [
            "Attention",
            "FeedForward"
        ],
        "module_algo_map": {
            "Attention": {
                "factor": 12
            },
            "FeedForward": {
                "factor": 6
            }
        }
    }
}

Datasets

GEN13-COVERS-REV03-512

• Repeats: 1
• Total number of images: 30
• Total number of aspect buckets: 5
• Resolution: 0.262144 megapixels
• Cropped: True
• Crop style: random
• Crop aspect: closest

GEN13-COVERS-REV03-768

• Repeats: 1
• Total number of images: 30
• Total number of aspect buckets: 1
• Resolution: 0.589824 megapixels
• Cropped: True
• Crop style: random
• Crop aspect: closest

GEN13-COVERS-REV03-1024

• Repeats: 1
• Total number of images: 30
• Total number of aspect buckets: 5
• Resolution: 1.048576 megapixels
• Cropped: True
• Crop style: random
• Crop aspect: closest

GEN13-PANELS-REV03-512

• Repeats: 0
• Total number of images: 176
• Total number of aspect buckets: 2
• Resolution: 0.262144 megapixels
• Cropped: True
• Crop style: random
• Crop aspect: closest

GEN13-PANELS-REV03-768

• Repeats: 0
• Total number of images: 175
• Total number of aspect buckets: 1
• Resolution: 0.589824 megapixels
• Cropped: True
• Crop style: random
• Crop aspect: closest

GEN13-PANELS-REV03-1024

• Repeats: 0
• Total number of images: 176
• Total number of aspect buckets: 2
• Resolution: 1.048576 megapixels
• Cropped: True
• Crop style: random
• Crop aspect: closest

GEN13-FRAGMENTS-REV03-512

• Repeats: 1
• Total number of images: 30
• Total number of aspect buckets: 16
• Resolution: 0.262144 megapixels
• Cropped: True
• Crop style: random
• Crop aspect: closest

GEN13-FRAGMENTS-REV03-768

• Repeats: 1
• Total number of images: 108
• Total number of aspect buckets: 30
• Resolution: 0.589824 megapixels
• Cropped: True
• Crop style: random
• Crop aspect: closest

GEN13-FRAGMENTS-REV03-1024

• Repeats: 1
• Total number of images: 110
• Total number of aspect buckets: 26
• Resolution: 1.048576 megapixels
• Cropped: True
• Crop style: random
• Crop aspect: closest

Inference

import torch
from diffusers import DiffusionPipeline
from lycoris import create_lycoris_from_weights

model_id = 'black-forest-labs/FLUX.1-dev'
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 = "caitlin and roxy sit together on a rooftop under the setting sun, the golden light bathing their faces in a soft glow, caitlin tall, athletic frame is relaxed as she leans back, dressed in a casual tank top that highlights her muscular build, while roxy, much smaller in stature, is perched next to her in a loose-fitting hoodie, her pink hair tousled by the wind, they share a quiet laugh, their bond of friendship bringing out a sense of relaxed femininity in both of them, contrasting their tough exteriors"

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=768,
    height=1344,
    guidance_scale=3.5,
).images[0]
image.save("output.png", format="PNG")