gattaplayer/besch-style-st-sd3.5-lora-adamw-1e-5-bs2-v01-1050

besch-style-st-sd3.5-lora-adamw-1e-5-bs2-v01-1050

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

No validation prompt was used during training.

None

Validation settings

• —CFG: 5.5
• —CFG Rescale: 0.0
• —Steps: 20
• —Sampler: None
• —Seed: 42
• —Resolution: 832x1216

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: 46
• —Training steps: 10500
• —Learning rate: 1e-05
• —Max grad norm: 0.01
• —Effective batch size: 2
• —Micro-batch size: 2
• —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: 64
• —LoRA Alpha: 64.0
• —LoRA Dropout: 0.1
• —LoRA initialisation style: default

Datasets

BESCH-SD35-V01-512

• —Repeats: 1
• —Total number of images: 70
• —Total number of aspect buckets: 6
• —Resolution: 0.262144 megapixels
• —Cropped: True
• —Crop style: random
• —Crop aspect: closest
• —Used for regularisation data: No

BESCH-SD35-V01-768

• —Repeats: 1
• —Total number of images: 70
• —Total number of aspect buckets: 6
• —Resolution: 0.589824 megapixels
• —Cropped: True
• —Crop style: random
• —Crop aspect: closest
• —Used for regularisation data: No

BESCH-SD35-V01-1024

• —Repeats: 1
• —Total number of images: 70
• —Total number of aspect buckets: 6
• —Resolution: 1.048576 megapixels
• —Cropped: True
• —Crop style: random
• —Crop aspect: closest
• —Used for regularisation data: No

Inference

import torch
from diffusers import DiffusionPipeline

model_id = 'stabilityai/stable-diffusion-3.5-large'
adapter_id = 'gattaplayer/besch-style-st-sd3.5-lora-adamw-1e-5-bs2-v01-1050'
pipeline = DiffusionPipeline.from_pretrained(model_id)
pipeline.load_lora_weights(adapter_id)

prompt = "An astronaut is riding a horse through the jungles of Thailand."
negative_prompt = 'blurry, cropped, ugly'
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=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=832,
    height=1216,
    guidance_scale=5.5,
).images[0]
image.save("output.png", format="PNG")