growwithdaisy/nwdxtrstrt_flat_20241217_163036

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

The main validation prompt used during training was:

a photo of a daisy

Validation settings

CFG: 3.5
CFG Rescale: 0.0
Steps: 28
Sampler: FlowMatchEulerDiscreteScheduler
Seed: 69
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
A room with two white chairs and a black table. There is a white door on the wall. There are two chairs in front of the door. There a black chair in front a white wall., The Henri Swivel Chair, Maiden Home Style

Negative Prompt
blurry, cropped, ugly

Prompt
A black table sits on top of four wooden legs. The table is shaped like an octagon. The floor under the table is gray., Studio Balestra Style

Negative Prompt
blurry, cropped, ugly

Prompt
Beetle Collection, A room with four chairs and a table in it. There are four chairs in front of the table. There is a rug under the chairs. There doors on the wall behind the chairs are blue., Gubi Style

Negative Prompt
blurry, cropped, ugly

Prompt
Satellite Collection, A black and white lamp is hanging on the wall. There is a white brick wall behind the lamp. There are white curtains on the window. There a black table in front of the wall with books on top of it., Gubi Style

Negative Prompt
blurry, cropped, ugly

Prompt
a photo of a daisy

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: 74
Training steps: 8000
Learning rate: 0.0002
- Learning rate schedule: constant
- Warmup steps: 0
Max grad norm: 2.0
Effective batch size: 8
- Micro-batch size: 2
- Gradient accumulation steps: 1
- Number of GPUs: 4
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: optimi-stableadamwweight_decay=1e-3
Trainable parameter precision: Pure BF16
Caption dropout probability: 5.0%

LyCORIS Config:

{
    "algo": "lokr",
    "multiplier": 1,
    "linear_dim": 1000000,
    "linear_alpha": 1,
    "factor": 16,
    "init_lokr_norm": 0.001,
    "apply_preset": {
        "target_module": [
            "FluxTransformerBlock",
            "FluxSingleTransformerBlock"
        ],
        "module_algo_map": {
            "Attention": {
                "factor": 16
            },
            "FeedForward": {
                "factor": 8
            }
        }
    }
}

Datasets

nwdxtrstrt_flat-512

Repeats: 0
Total number of images: ~284
Total number of aspect buckets: 1
Resolution: 0.262144 megapixels
Cropped: False
Crop style: None
Crop aspect: None
Used for regularisation data: No

nwdxtrstrt_flat-768

Repeats: 0
Total number of images: ~284
Total number of aspect buckets: 4
Resolution: 0.589824 megapixels
Cropped: False
Crop style: None
Crop aspect: None
Used for regularisation data: No

nwdxtrstrt_flat-1024

Repeats: 0
Total number of images: ~240
Total number of aspect buckets: 2
Resolution: 1.048576 megapixels
Cropped: False
Crop style: None
Crop aspect: None
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 = 'playerzer0x/growwithdaisy/nwdxtrstrt_flat_20241217_163036'
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 = "a photo of a daisy"


## Optional: quantise the model to save on vram.
## Note: The model was not quantised during training, so it is not necessary to quantise it 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=28,
    generator=torch.Generator(device='cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu').manual_seed(69),
    width=1024,
    height=1024,
    guidance_scale=3.5,
).images[0]
image.save("output.png", format="PNG")

growwithdaisy
/

nwdxtrstrt_flat_20241217_163036