tabler-icon-generator

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

The main validation prompt used during training was:

minimalist icon

Validation settings

• CFG: 3.0
• CFG Rescale: 0.0
• Steps: 20
• Sampler: FlowMatchEulerDiscreteScheduler
• Seed: 42
• Resolution: 512x512
• 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

minimalist icon alert circle, black on white background, clean lines

Negative Prompt

blurry, cropped, ugly

Prompt

minimalist icon smiley face, black on white background, clean lines

Negative Prompt

blurry, cropped, ugly

Prompt

minimalist icon facebook logo, black on white background, clean lines

Negative Prompt

blurry, cropped, ugly

Prompt

minimalist icon hd badge, black on white background, clean lines

Negative Prompt

blurry, cropped, ugly

Prompt

minimalist icon coin off, black on white background, clean lines

Negative Prompt

blurry, cropped, ugly

Prompt

minimalist icon

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: 10000

• Learning rate: 0.0001

  • 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: 0.0%

• LoRA Rank: 16

• LoRA Alpha: None

• LoRA Dropout: 0.1

• LoRA initialisation style: default

Datasets

tabler-icons-512

• Repeats: 0
• Total number of images: 4739
• Total number of aspect buckets: 1
• Resolution: 0.262144 megapixels
• Cropped: False
• Crop style: None
• Crop aspect: None
• Used for regularisation data: No

Inference

import torch
from diffusers import DiffusionPipeline

model_id = 'black-forest-labs/FLUX.1-dev'
adapter_id = 'noahyoungs/tabler-icon-generator'
pipeline = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.bfloat16) # loading directly in bf16
pipeline.load_lora_weights(adapter_id)

prompt = "minimalist icon"


## 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=512,
    height=512,
    guidance_scale=3.0,
).images[0]
image.save("output.png", format="PNG")