subject-lora

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

The main validation prompt used during training was:

This image shows a y3o golden retriever happily standing on a sunlit Caribbean beach, its mouth slightly open, revealing a playful pink tongue. The dog's thick, lustrous golden coat, characteristic of the breed, shines warmly under the tropical sun, while its gentle, dark eyes reflect a joyful and friendly demeanor. Behind the retriever, crystal-clear turquoise waters gently lap against pristine white sand, and lush palm trees sway softly in the ocean breeze, creating an idyllic Caribbean setting that perfectly complements the dog's cheerful presence.

Validation settings

CFG: 7.5
CFG Rescale: 0.0
Steps: 30
Sampler: None
Seed: 42
Resolution: 1024

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
'

Prompt
This image shows a y3o golden retriever happily standing on a sunlit Caribbean beach, its mouth slightly open, revealing a playful pink tongue. The dog's thick, lustrous golden coat, characteristic of the breed, shines warmly under the tropical sun, while its gentle, dark eyes reflect a joyful and friendly demeanor. Behind the retriever, crystal-clear turquoise waters gently lap against pristine white sand, and lush palm trees sway softly in the ocean breeze, creating an idyllic Caribbean setting that perfectly complements the dog's cheerful presence.

Negative Prompt
'

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

Training settings

Training epochs: 49
Training steps: 300
Learning rate: 0.000505
Max grad norm: 0.01
Effective batch size: 12
- Micro-batch size: 12
- Gradient accumulation steps: 1
- Number of GPUs: 1
Prediction type: flow-matching
Rescaled betas zero SNR: False
Optimizer: adamw_bf16
Precision: Pure BF16
Quantised: No
Xformers: Not used
LoRA Rank: 768
LoRA Alpha: 768.0
LoRA Dropout: 0.1
LoRA initialisation style: default

Datasets

subject

Repeats: 5
Total number of images: 6
Total number of aspect buckets: 1
Resolution: 1.0 megapixels
Cropped: False
Crop style: None
Crop aspect: None
Used for regularisation data: No

Inference

import torch
from diffusers import DiffusionPipeline

model_id = 'stabilityai/stable-diffusion-3.5-large'
adapter_id = 'AngelZeur/subject-lora'
pipeline = DiffusionPipeline.from_pretrained(model_id)
pipeline.load_lora_weights(adapter_id)

prompt = "This image shows a y3o golden retriever happily standing on a sunlit Caribbean beach, its mouth slightly open, revealing a playful pink tongue. The dog's thick, lustrous golden coat, characteristic of the breed, shines warmly under the tropical sun, while its gentle, dark eyes reflect a joyful and friendly demeanor. Behind the retriever, crystal-clear turquoise waters gently lap against pristine white sand, and lush palm trees sway softly in the ocean breeze, creating an idyllic Caribbean setting that perfectly complements the dog's cheerful presence."
negative_prompt = ''
pipeline.to('cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu')
image = pipeline(
    prompt=prompt,
    negative_prompt=negative_prompt,
    num_inference_steps=30,
    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=7.5,
).images[0]
image.save("output.png", format="PNG")