growwithdaisy/mnmlsmo_wearables_20241104_163505

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

The main validation prompt used during training was:

photo of a daisy mnmlsmo wearable

Validation settings

• CFG: 3.5
• CFG Rescale: 0.0
• Steps: 20
• Sampler: None
• Seed: 69
• Resolution: 1024x1024

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

mnmlsmo wearables, The image is a close-up of a wristwatch with a black strap. The watch has a square-shaped face with a silver-colored metal case and a black dial. The face of the watch is white with black hands and markers, and there is a small second hand at the 3 o'clock position.

Negative Prompt

blurry, cropped, ugly

Prompt

mnmlsmo wearables, The image is a digital illustration of a wristwatch with a black strap. The watch face is square in shape and has a white background with black numbers and hands. The hands of the watch are black and the numbers are in a modern, minimalist design.

Negative Prompt

blurry, cropped, ugly

Prompt

mnmlsmo wearables, The image is a close-up of a wristwatch with a square-shaped face. The watch has a black strap with a silver buckle on the right side. The face of the watch is white with black numbers and hands, and there is a small black hour hand at the 3 o'clock position.

Negative Prompt

blurry, cropped, ugly

Prompt

mnmlsmo wearables, The image shows an aerial view of a large swimming pool on a rocky cliff overlooking the ocean. The pool is rectangular in shape and has a blue-green color. There are several people in the pool, some of whom are swimming and some are sitting on the beach. The water is crystal clear and the rocks around the pool are jagged and rocky.

Negative Prompt

blurry, cropped, ugly

Prompt

photo of a daisy mnmlsmo wearable

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: 90
• Training steps: 1000
• Learning rate: 0.0004
• Max grad norm: 2.0
• Effective batch size: 8
  • Micro-batch size: 1
  • Gradient accumulation steps: 1
  • Number of GPUs: 8
• Prediction type: flow-matching (flux parameters=['shift=3', 'flux_guidance_value=1.0'])
• Rescaled betas zero SNR: False
• Optimizer: optimi-stableadamwweight_decay=1e-3
• Precision: Pure BF16
• Quantised: No
• Xformers: Not used
• 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

mnmlsmo_wearables-512

• Repeats: 0
• Total number of images: ~32
• Total number of aspect buckets: 1
• Resolution: 0.262144 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None
• Used for regularisation data: No

mnmlsmo_wearables-768

• Repeats: 0
• Total number of images: ~32
• Total number of aspect buckets: 3
• Resolution: 0.589824 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None
• Used for regularisation data: No

mnmlsmo_wearables-1024

• Repeats: 0
• Total number of images: ~24
• 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

model_id = 'black-forest-labs/FLUX.1-dev'
adapter_id = 'pytorch_lora_weights.safetensors' # you will have to download this manually
lora_scale = 1.0
wrapper, _ = create_lycoris_from_weights(lora_scale, adapter_id, pipeline.transformer)
wrapper.merge_to()

prompt = "photo of a daisy mnmlsmo wearable"

pipeline.to('cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu')
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(1641421826),
    width=1024,
    height=1024,
    guidance_scale=3.5,
).images[0]
image.save("output.png", format="PNG")