home-interior-4

home-interior-4

This is a standard PEFT LoRA derived from stabilityai/stable-diffusion-3.5-large.

The main validation prompt used during training was:

home-interior. two brown chairs and a coffee table. A framed abstract artwork hangs above the armchairs. The walls are painted in a soft neutral tone, and a large flat-screen TV is mounted on wall. A rug covers part of the light oak flooring. On the left, large black-framed glass sliding doors lead to outdoor.

Validation settings

• CFG: 3.0
• CFG Rescale: 0.0
• Steps: 25
• Sampler: FlowMatchEulerDiscreteScheduler
• Seed: 42
• Resolutions: 1024x1024, 1024x896, 1280x768
• Skip-layer guidance: skip_guidance_layers=[7, 8, 9],

Note: The validation settings are not necessarily the same as the training settings.

You can find some example images in the following gallery:

<Gallery />

The text encoder was not trained. You may reuse the base model text encoder for inference.

Training settings

• Training epochs: 9

• Training steps: 2000

• Learning rate: 1e-05

  • Learning rate schedule: polynomial
  • Warmup steps: 100

• Max grad norm: 0.1

• Effective batch size: 4

  • Micro-batch size: 4
  • Gradient accumulation steps: 1
  • Number of GPUs: 1

• Gradient checkpointing: True

• Prediction type: flow-matching (extra parameters=['shift=1.0'])

• Optimizer: adamw_bf16

• Trainable parameter precision: FP32

• Caption dropout probability: 0.0%

• LoRA Rank: 128

• LoRA Alpha: None

• LoRA Dropout: 0.1

• LoRA initialisation style: default

Datasets

home-interior-4

• Repeats: 3
• Total number of images: 209
• Total number of aspect buckets: 9
• Resolution: 1.048576 megapixels
• Cropped: True
• Crop style: random
• Crop aspect: preserve
• Used for regularisation data: No

Inference

import torch
from diffusers import DiffusionPipeline

model_id = 'stabilityai/stable-diffusion-3.5-large'
adapter_id = 'daehuncho/home-interior-4'
pipeline = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.bfloat16) # loading directly in bf16
pipeline.load_lora_weights(adapter_id)

prompt = "home-interior. two brown chairs and a coffee table. A framed abstract artwork hangs above the armchairs. The walls are painted in a soft neutral tone, and a large flat-screen TV is mounted on wall. A rug covers part of the light oak flooring. On the left, large black-framed glass sliding doors lead to outdoor."
negative_prompt = 'blurry, cropped, ugly'

## 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,
    negative_prompt=negative_prompt,
    num_inference_steps=25,
    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,
    skip_guidance_layers=[7, 8, 9],
).images[0]
image.save("output.png", format="PNG")