rangerover-phase1-2e-4-constant

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

No validation prompt was used during training.

None

Validation settings

CFG: 3.0
CFG Rescale: 0.0
Steps: 20
Sampler: FlowMatchEulerDiscreteScheduler
Seed: 42
Resolution: 1344x768
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
A Range Rover driving down the highway of Hong Kong, the beautiful cityscape is behind it. It is night time and the night lights of the skyscrapers are dazzling.

Negative Prompt
blurry, cropped, ugly

Prompt
A pink Range Rover

Negative Prompt
blurry, cropped, ugly

Prompt
Create a breathtaking, high-quality commercial-style image of a rugged yet luxurious Range Rover navigating a serene, misty forest. The vehicle should be the focal point, showcasing its robust design, matte finish, and adventurous spirit. The forest should be dense with towering evergreen trees, soft sunlight filtering through the canopy, and a faint mist lingering in the air. The Range Rover is parked on a muddy trail, with its tires slightly sunk into the earth, emphasizing its off-road capability. In the background, a crystal-clear river winds through the trees, reflecting the soft light. Add a sense of tranquility and adventure, with a tagline in elegant, nature-inspired typography at the bottom: 'Range Rover: Conquer the Wild in Style.' The overall mood should be immersive, bold, and inspiring, appealing to adventurers and luxury enthusiasts alike.

Negative Prompt
blurry, cropped, ugly

Prompt
The Range Rover is positioned on a flat, light-colored surface against a gradient background that transitions from a pale yellow to a light blue. The lighting is soft and diffused, creating subtle shadows beneath the Range Rover and highlighting its metallic paint with gentle reflections. The Range Rover is shown from a side profile perspective, with the entire vehicle visible within the frame.

Negative Prompt
blurry, cropped, ugly

Prompt
A Range Rover.

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: 2
Training steps: 7250
Learning rate: 0.0002
- Learning rate schedule: constant
- 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

klimt-background-512

Repeats: 22
Total number of images: 69
Total number of aspect buckets: 11
Resolution: 0.262144 megapixels
Cropped: False
Crop style: None
Crop aspect: None
Used for regularisation data: No

klimt-background-768

Repeats: 22
Total number of images: 69
Total number of aspect buckets: 1
Resolution: 0.589824 megapixels
Cropped: False
Crop style: None
Crop aspect: None
Used for regularisation data: No

klimt-background-1024

Repeats: 11
Total number of images: 69
Total number of aspect buckets: 15
Resolution: 1.048576 megapixels
Cropped: False
Crop style: None
Crop aspect: None
Used for regularisation data: No

klimt-background-1536

Repeats: 5
Total number of images: 68
Total number of aspect buckets: 19
Resolution: 2.359296 megapixels
Cropped: False
Crop style: None
Crop aspect: None
Used for regularisation data: No

klimt-background-512-crop

Repeats: 11
Total number of images: 69
Total number of aspect buckets: 1
Resolution: 0.262144 megapixels
Cropped: True
Crop style: random
Crop aspect: square
Used for regularisation data: No

klimt-background-768-crop

Repeats: 11
Total number of images: 69
Total number of aspect buckets: 1
Resolution: 0.589824 megapixels
Cropped: True
Crop style: random
Crop aspect: square
Used for regularisation data: No

klimt-background-512-tight-crop

Repeats: 11
Total number of images: 69
Total number of aspect buckets: 1
Resolution: 0.262144 megapixels
Cropped: True
Crop style: random
Crop aspect: square
Used for regularisation data: No

klimt-background-768-tight-crop

Repeats: 11
Total number of images: 69
Total number of aspect buckets: 1
Resolution: 0.589824 megapixels
Cropped: True
Crop style: random
Crop aspect: square
Used for regularisation data: No

klimt-background-1024-crop

Repeats: 5
Total number of images: 68
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/rangerover-phase1-2e-4-constant'
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=1344,
    height=768,
    guidance_scale=3.0,
).images[0]
image.save("output.png", format="PNG")

mipat12
/

rangerover-phase1-2e-4-constant