quirky-product-render-lora

This is a standard PEFT LoRA derived from black-forest-labs/FLUX.1-dev.

The main validation prompt used during training was:

a refined product render of a tennis racket in the style of an Apple product. This evokes throwback Y2K vibes, but still retains the sleek elegance of an Apple product. The racket is made out of a metallic silver and has a glowing sensor on it.

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

a refined product render of a tennis racket in the style of an Apple product. This evokes throwback Y2K vibes, but still retains the sleek elegance of an Apple product. The racket is made out of a metallic silver and has a glowing sensor on it.

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: 2000

• Learning rate: 0.0003

  • Learning rate schedule: polynomial
  • Warmup steps: 100

• Max grad norm: 2.0

• 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', 'flux_lora_target=all'])

• Optimizer: adamw_bf16

• Trainable parameter precision: Pure BF16

• Caption dropout probability: 5.0%

• LoRA Rank: 1

• LoRA Alpha: None

• LoRA Dropout: 0.1

• LoRA initialisation style: default

Datasets

my-quirky-product-dataset-256

• Repeats: 200
• Total number of images: 9
• Total number of aspect buckets: 2
• Resolution: 0.065536 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None
• Used for regularisation data: No

my-quirky-product-dataset-crop-256

• Repeats: 200
• Total number of images: 9
• Total number of aspect buckets: 1
• Resolution: 0.065536 megapixels
• Cropped: True
• Crop style: center
• Crop aspect: square
• Used for regularisation data: No

my-quirky-product-dataset-512

• Repeats: 200
• Total number of images: 9
• Total number of aspect buckets: 1
• Resolution: 0.262144 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None
• Used for regularisation data: No

my-quirky-product-dataset-crop-512

• Repeats: 200
• Total number of images: 9
• Total number of aspect buckets: 1
• Resolution: 0.262144 megapixels
• Cropped: True
• Crop style: center
• Crop aspect: square
• Used for regularisation data: No

my-quirky-product-dataset-768

• Repeats: 200
• Total number of images: 9
• Total number of aspect buckets: 1
• Resolution: 0.589824 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None
• Used for regularisation data: No

my-quirky-product-dataset-crop-768

• Repeats: 200
• Total number of images: 9
• Total number of aspect buckets: 1
• Resolution: 0.589824 megapixels
• Cropped: True
• Crop style: center
• Crop aspect: square
• Used for regularisation data: No

my-quirky-product-dataset-1024

• Repeats: 200
• Total number of images: 9
• Total number of aspect buckets: 3
• Resolution: 1.048576 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None
• Used for regularisation data: No

my-quirky-product-dataset-crop-1024

• Repeats: 200
• Total number of images: 9
• Total number of aspect buckets: 1
• Resolution: 1.048576 megapixels
• Cropped: True
• Crop style: center
• Crop aspect: square
• Used for regularisation data: No

my-quirky-product-dataset-1440

• Repeats: 200
• Total number of images: 9
• Total number of aspect buckets: 3
• Resolution: 2.0736 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None
• Used for regularisation data: No

my-quirky-product-dataset-crop-1440

• Repeats: 200
• Total number of images: 7
• Total number of aspect buckets: 1
• Resolution: 2.0736 megapixels
• Cropped: True
• Crop style: center
• Crop aspect: square
• Used for regularisation data: No

Inference

import torch
from diffusers import DiffusionPipeline

model_id = 'black-forest-labs/FLUX.1-dev'
adapter_id = 'fzbuzz/quirky-product-render-lora'
pipeline = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.bfloat16) # loading directly in bf16
pipeline.load_lora_weights(adapter_id)

prompt = "a refined product render of a tennis racket in the style of an Apple product. This evokes throwback Y2K vibes, but still retains the sleek elegance of an Apple product. The racket is made out of a metallic silver and has a glowing sensor on it."


## 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")