simpletuner-lora

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

The main validation prompt used during training was:

windows test 6 Treatment Lotion being dispensed. Its silky, translucent texture glistens under a soft, diffused light against a minimalist white background with subtle ocean ripple effects

Validation settings

CFG: 3.0
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
The La Mer Treatment Lotion bottle rests on a polished glass surface with a glowing mist swirling around it. Soft oceanic hues of deep blue and aquamarine radiate in the background, creating an ethereal and luxurious atmosphere.

Negative Prompt
blurry, cropped, ugly

Prompt
A macro shot of the La Mer Treatment Lotion being dispensed. Its silky, translucent texture glistens under soft, diffused light against a minimalist white background with subtle ocean ripple effects.

Negative Prompt
blurry, cropped, ugly

Prompt
The La Mer Treatment Lotion bottle is placed on a smooth coastal rock surrounded by fresh sea kelp and gently flowing water. The sunlight filters through a soft morning haze, evoking natural purity and serenity.

Negative Prompt
blurry, cropped, ugly

Prompt
The La Mer Treatment Lotion bottle is set against a minimalist, abstract backdrop of a glowing world map. Subtle ocean-inspired hues and soft lighting emphasize universal luxury and timeless appeal.

Negative Prompt
blurry, cropped, ugly

Prompt
windows test 6 Treatment Lotion being dispensed. Its silky, translucent texture glistens under a soft, diffused light against a minimalist white background with subtle ocean ripple effects

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: 0
Training steps: 15
Learning rate: 0.0001
- Learning rate schedule: polynomial
- Warmup steps: 100
Max grad norm: 1.0
Effective batch size: 1
- Micro-batch size: 1
- Gradient accumulation steps: 1
- Number of GPUs: 1
Gradient checkpointing: True
Prediction type: flow-matching (extra parameters=['shift=3', '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

La_Mer_Treatment_Lotion_FLUX_v01

Repeats: 1000
Total number of images: 34
Total number of aspect buckets: 1
Resolution: 1.048576 megapixels
Cropped: True
Crop style: center
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 = 'salmonrusty/simpletuner-lora'
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 = "windows test 6 Treatment Lotion being dispensed. Its silky, translucent texture glistens under a soft, diffused light against a minimalist white background with subtle ocean ripple effects"


## 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=3.0,
).images[0]
image.save("output.png", format="PNG")

salmonrusty
/

simpletuner-lora