Hugging Face's logo

mipat12
/
michelangelo-phase1-4e-4-ss3.0

Text-to-Image
Diffusers
Safetensors
flux
flux-diffusers
simpletuner
Not-For-All-Audiences
lora
template:sd-lora
lycoris
Model card Files Community

michelangelo-phase1-4e-4-ss3.0

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

No validation prompt was used during training.

None

Validation settings

  • CFG: 2.5
  • CFG Rescale: 0.0
  • Steps: 20
  • Sampler: FlowMatchEulerDiscreteScheduler
  • Seed: 42
  • Resolution: 1024x1024
  • Skip-layer guidance:

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
in the style of a Michelangelo sculpture, Three elderly women huddle together, their robes intertwined as they share a scroll between them. Their faces show deep concentration, with pronounced wrinkles and hollow cheeks.
Negative Prompt
blurry, cropped, ugly
Prompt
in the style of a Michelangelo sculpture, a hamster
Negative Prompt
blurry, cropped, ugly
Prompt
in the style of a Michelangelo sculpture, A plump hamster sits upright on its haunches, tiny paws clutching a seed with remarkable dignity. Its fur is rendered in detailed marble ripples, while its alert ears are tilted forward attentively. The creature's round cheeks suggest stored food, and its whiskers are delicately carved. The base is decorated with miniature carved leaves and fallen seeds, while the background remains unadorned and shadowed.
Negative Prompt
blurry, cropped, ugly
Prompt
in the style of a Michelangelo sculpture, A Range Rover emerges from solid marble, its commanding presence emphasized by strong angular lines and bold proportions. The vehicle rests in a three-quarter pose, with its distinctive grille and headlights carved in meticulous detail. Each wheel arch suggests latent motion, while the smooth curves of the hood flow into the upright windscreen. The base appears to ripple like terrain beneath the wheels, suggesting the vehicle's adventurous nature. The background is stark, drawing attention to the interplay of light and shadow across the sculptured surfaces.
Negative Prompt
blurry, cropped, ugly
Prompt
in the style of a Michelangelo sculpture, A young girl stands on tiptoes reaching upward, her hair falling in loose waves. A ribbon streams behind her, caught in an invisible wind. The base beneath her feet shows carved clouds, suggesting she floats between earth and sky.
Negative Prompt
blurry, cropped, ugly
Prompt
a man holding a sign that says, 'this is a sign
Negative Prompt
blurry, cropped, ugly
Prompt
a pig, in a post apocalyptic world, with a shotgun, in a leather jacket, in a desert, with a motorcycle
Negative Prompt
blurry, cropped, ugly
Prompt
in the style of a Michelangelo sculpture, woman holding a sign that says 'I LOVE PROMPTS!'.
Negative Prompt
blurry, cropped, ugly
Prompt
in the style of a Michelangelo sculpture, A nude male figure stands tall on a pedestal, his left arm is raised, while his right arm hangs freely by his side. The figure has curly hair and a focused, determined expression on his face looking slightly to his left. In the background, there are large panels with rectangular details on the walls, suggesting an indoor setting.
Negative Prompt
blurry, cropped, ugly
Prompt
in the style of a Michelangelo sculpture, a bearded man sits down dressed in long garments. The background is plain.
Negative Prompt
blurry, cropped, ugly
Prompt
in the style of a Michelangelo sculpture, a woman cradling a lifeless man on her lap. The woman wears draped clothing with a hood, and has a sorrowful expression. The man is depicted naked except for a cloth around his waist, his arms and legs extended lifelessly. The background is an intricate marble wall with a mixed pattern of colors (brown, green, beige).
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: 15
  • Training steps: 5000
  • Learning rate: 0.0004
    • Learning rate schedule: polynomial
    • Warmup steps: 100
  • Max grad norm: 0.1
  • Effective batch size: 3
    • Micro-batch size: 3
    • Gradient accumulation steps: 1
    • Number of GPUs: 1
  • Gradient checkpointing: True
  • Prediction type: flow-matching (extra parameters=['shift=3.0', 'flux_guidance_mode=constant', 'flux_guidance_value=1.0', 'flow_matching_loss=compatible'])
  • Optimizer: adamw_bf16
  • Trainable parameter precision: Pure BF16
  • Caption dropout probability: 10.0%

LyCORIS Config: 

{
    "algo": "lokr",
    "multiplier": 1.0,
    "linear_dim": 10000,
    "linear_alpha": 1,
    "factor": 16,
    "apply_preset": {
        "target_module": [
            "Attention",
            "FeedForward"
        ],
        "module_algo_map": {
            "Attention": {
                "factor": 16
            },
            "FeedForward": {
                "factor": 8
            }
        }
    }
}

Datasets

davinci-512

  • Repeats: 11
  • Total number of images: 13
  • Total number of aspect buckets: 7
  • Resolution: 0.262144 megapixels
  • Cropped: False
  • Crop style: None
  • Crop aspect: None
  • Used for regularisation data: No

davinci-768

  • Repeats: 8
  • Total number of images: 13
  • Total number of aspect buckets: 8
  • Resolution: 0.589824 megapixels
  • Cropped: False
  • Crop style: None
  • Crop aspect: None
  • Used for regularisation data: No

davinci-1024

  • Repeats: 5
  • Total number of images: 13
  • Total number of aspect buckets: 10
  • Resolution: 1.048576 megapixels
  • Cropped: False
  • Crop style: None
  • Crop aspect: None
  • Used for regularisation data: No

davinci-1536

  • Repeats: 2
  • Total number of images: 13
  • Total number of aspect buckets: 10
  • Resolution: 2.359296 megapixels
  • Cropped: False
  • Crop style: None
  • Crop aspect: None
  • Used for regularisation data: No

davinci-crops-512

  • Repeats: 8
  • Total number of images: 13
  • Total number of aspect buckets: 1
  • Resolution: 0.262144 megapixels
  • Cropped: True
  • Crop style: random
  • Crop aspect: square
  • Used for regularisation data: No

davinci-1024-crop

  • Repeats: 5
  • Total number of images: 13
  • Total number of aspect buckets: 1
  • Resolution: 1.048576 megapixels
  • Cropped: True
  • Crop style: random
  • Crop aspect: square
  • Used for regularisation data: No

Inference 

import torch
from diffusers import DiffusionPipeline
from lycoris import create_lycoris_from_weights


def download_adapter(repo_id: str):
    import os
    from huggingface_hub import hf_hub_download
    adapter_filename = "pytorch_lora_weights.safetensors"
    cache_dir = os.environ.get('HF_PATH', os.path.expanduser('~/.cache/huggingface/hub/models'))
    cleaned_adapter_path = repo_id.replace("/", "_").replace("\\", "_").replace(":", "_")
    path_to_adapter = os.path.join(cache_dir, cleaned_adapter_path)
    path_to_adapter_file = os.path.join(path_to_adapter, adapter_filename)
    os.makedirs(path_to_adapter, exist_ok=True)
    hf_hub_download(
        repo_id=repo_id, filename=adapter_filename, local_dir=path_to_adapter
    )

    return path_to_adapter_file
    
model_id = 'black-forest-labs/FLUX.1-dev'
adapter_repo_id = 'mipat12/michelangelo-phase1-4e-4-ss3.0'
adapter_filename = 'pytorch_lora_weights.safetensors'
adapter_file_path = download_adapter(repo_id=adapter_repo_id)
pipeline = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.bfloat16) # loading directly in bf16
lora_scale = 1.0
wrapper, _ = create_lycoris_from_weights(lora_scale, adapter_file_path, pipeline.transformer)
wrapper.merge_to()

prompt = "An astronaut is riding a horse through the jungles of Thailand."


## Optional: quantise the model to save on vram.
## Note: The model was quantised during training, and so it is recommended to do the same during inference time.
from optimum.quanto import quantize, freeze, qint8
quantize(pipeline.transformer, weights=qint8)
freeze(pipeline.transformer)
    
pipeline.to('cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu') # the pipeline is already in its target precision level
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(42),
    width=1024,
    height=1024,
    guidance_scale=2.5,
).images[0]
image.save("output.png", format="PNG")

Exponential Moving Average (EMA)

SimpleTuner generates a safetensors variant of the EMA weights and a pt file.

The safetensors file is intended to be used for inference, and the pt file is for continuing finetuning.

The EMA model may provide a more well-rounded result, but typically will feel undertrained compared to the full model as it is a running decayed average of the model weights.

Downloads last month
308
Inference API
Text-to-Image
Examples

Model tree for mipat12/michelangelo-phase1-4e-4-ss3.0

Base model

black-forest-labs/FLUX.1-dev
Adapter
(13772)
this model