growwithdaisy/crtdt_20241209_192533_20241210_171640

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

The main validation prompt used during training was:

a man wearing a reverse white crtdt hot swingers tee

Validation settings

CFG: 3.5
CFG Rescale: 0.0
Steps: 20
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
crumpled green crtdt hitters only tee, on a white sheet

Negative Prompt
blurry, cropped, ugly

Prompt
a light-skinned woman, holding reverse green crtdt hitters only tee, wearing black clothes, standing on top of an area rug, a projector screen in the background with windows on walls, woman is looking at the tee, two art stands on either side of the woman with sculptures on them

Negative Prompt
blurry, cropped, ugly

Prompt
crumpled navy crtdt dont settle tee, on a white sheet

Negative Prompt
blurry, cropped, ugly

Prompt
the back of a tan-skinned man, wearing reverse white crtdt hot swingers tee and white sneakers and dark blue shorts, a light-skinned woman, wearing black clothes, right half of the room is dark and left half is lit up, green crtdt hitters only tee draped over the couch

Negative Prompt
blurry, cropped, ugly

Prompt
a man wearing a reverse white crtdt hot swingers tee

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: 166
Training steps: 2500
Learning rate: 0.0002
- Learning rate schedule: constant
- Warmup steps: 0
Max grad norm: 2.0
Effective batch size: 16
- Micro-batch size: 2
- Gradient accumulation steps: 1
- Number of GPUs: 8
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": 12,
    "init_lokr_norm": 0.001,
    "apply_preset": {
        "target_module": [
            "FluxTransformerBlock",
            "FluxSingleTransformerBlock"
        ],
        "module_algo_map": {
            "Attention": {
                "factor": 12
            },
            "FeedForward": {
                "factor": 6
            }
        }
    }
}

Datasets

crtdt_20241209_192533-512

Repeats: 0
Total number of images: ~56
Total number of aspect buckets: 4
Resolution: 0.262144 megapixels
Cropped: False
Crop style: None
Crop aspect: None
Used for regularisation data: No

crtdt_20241209_192533-768

Repeats: 0
Total number of images: ~56
Total number of aspect buckets: 4
Resolution: 0.589824 megapixels
Cropped: False
Crop style: None
Crop aspect: None
Used for regularisation data: No

crtdt_20241209_192533-1024

Repeats: 0
Total number of images: ~56
Total number of aspect buckets: 4
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/crtdt_20241209_192533_20241210_171640'
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 man wearing a reverse white crtdt hot swingers tee"


## 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=20,
    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
/

crtdt_20241209_192533_20241210_171640