sdxl-irina-v1

This is a LyCORIS adapter derived from stabilityai/stable-diffusion-xl-base-1.0.

No validation prompt was used during training.

None

Validation settings

CFG: 4.2
CFG Rescale: 0.0
Steps: 24
Sampler: ddim
Seed: 42
Resolution: 1024x1024

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
1r1n4, brown_eyes, black_hair, flat_chest, cyberpunk, futuristic_city, outdoor, neon_jacket, open_jacket, white_jeans, futuristic, cinematic

Negative Prompt
blurry, cropped, ugly

Prompt
1r1n4, brown_eyes, black_hair, flat_chest, fantasy, castle, outdoor, princess_dress, princess_tiara, crown, white_dress

Negative Prompt
blurry, cropped, ugly

Prompt
1r1n4, brown_eyes, black_hair, flat_chest, deepthroat, deep_throat, fellatio, irrumatio, throat_bulge, eyes_wide_open, smeared_makeup, huge_penis

Negative Prompt
blurry, cropped, ugly

Prompt
1r1n4, brown_eyes, black_hair, flat_chest, microskirt, croptop, pussy_peek, covered_nipples, posing, serious_face, provocative

Negative Prompt
blurry, cropped, ugly

Prompt
1r1n4, brown_eyes, black_hair, bob_cut, huge_breasts, cleavage, bodysuit, tight_clothing, undersized_clothing

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: 20
Training steps: 20000
Learning rate: 0.0001
- Learning rate schedule: polynomial
- Warmup steps: 100
Max grad value: 2.0
Effective batch size: 3
- Micro-batch size: 3
- Gradient accumulation steps: 1
- Number of GPUs: 1
Gradient checkpointing: True
Prediction type: epsilon (extra parameters=['training_scheduler_timestep_spacing=trailing', 'inference_scheduler_timestep_spacing=trailing'])
Optimizer: adamw_bf16
Trainable parameter precision: Pure BF16
Base model precision: no_change
Caption dropout probability: 0.05%

LyCORIS Config:

{
    "algo": "lokr",
    "multiplier": 1.0,
    "linear_dim": 10000,
    "linear_alpha": 1,
    "factor": 16,
    "apply_preset": {
        "target_module": [
            "Attention",
            "FeedForward"
        ],
        "module_algo_map": {
            "Attention": {
                "factor": 16
            },
            "FeedForward": {
                "factor": 8
            }
        }
    }
}

Datasets

irina-dataset-512

Repeats: 10
Total number of images: 59
Total number of aspect buckets: 9
Resolution: 0.262144 megapixels
Cropped: False
Crop style: None
Crop aspect: None
Used for regularisation data: No

irina-dataset-1024

Repeats: 10
Total number of images: 59
Total number of aspect buckets: 2
Resolution: 1.048576 megapixels
Cropped: False
Crop style: None
Crop aspect: None
Used for regularisation data: No

irina-dataset-512-crop

Repeats: 10
Total number of images: 59
Total number of aspect buckets: 1
Resolution: 0.262144 megapixels
Cropped: True
Crop style: random
Crop aspect: square
Used for regularisation data: No

irina-dataset-1024-crop

Repeats: 10
Total number of images: 59
Total number of aspect buckets: 1
Resolution: 1.048576 megapixels
Cropped: True
Crop style: random
Crop aspect: square
Used for regularisation data: No

Inference

import torch
from diffusers import DiffusionPipeline
from lycoris import create_lycoris_from_weights


def download_adapter(repo_id: str):
    import os
    from huggingface_hub import hf_hub_download
    adapter_filename = "pytorch_lora_weights.safetensors"
    cache_dir = os.environ.get('HF_PATH', os.path.expanduser('~/.cache/huggingface/hub/models'))
    cleaned_adapter_path = repo_id.replace("/", "_").replace("\\", "_").replace(":", "_")
    path_to_adapter = os.path.join(cache_dir, cleaned_adapter_path)
    path_to_adapter_file = os.path.join(path_to_adapter, adapter_filename)
    os.makedirs(path_to_adapter, exist_ok=True)
    hf_hub_download(
        repo_id=repo_id, filename=adapter_filename, local_dir=path_to_adapter
    )

    return path_to_adapter_file
    
model_id = 'stabilityai/stable-diffusion-xl-base-1.0'
adapter_repo_id = 'Tripater/sdxl-irina-v1'
adapter_filename = 'pytorch_lora_weights.safetensors'
adapter_file_path = download_adapter(repo_id=adapter_repo_id)
pipeline = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.bfloat16) # loading directly in bf16
lora_scale = 1.0
wrapper, _ = create_lycoris_from_weights(lora_scale, adapter_file_path, pipeline.unet)
wrapper.merge_to()

prompt = "An astronaut is riding a horse through the jungles of Thailand."
negative_prompt = 'blurry, cropped, ugly'

## Optional: quantise the model to save on vram.
## Note: The model was not quantised during training, so it is not necessary to quantise it during inference time.
#from optimum.quanto import quantize, freeze, qint8
#quantize(pipeline.unet, weights=qint8)
#freeze(pipeline.unet)
    
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
model_output = pipeline(
    prompt=prompt,
    negative_prompt=negative_prompt,
    num_inference_steps=24,
    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=4.2,
    guidance_rescale=0.0,
).images[0]

model_output.save("output.png", format="PNG")

Tripater
/

sdxl-irina-v1