metadata
license: other
base_model: black-forest-labs/FLUX.1-dev
tags:
- flux
- flux-diffusers
- text-to-image
- image-to-image
- diffusers
- simpletuner
- not-for-all-audiences
- lora
- template:sd-lora
- lycoris
pipeline_tag: text-to-image
inference: true
widget:
- text: unconditional (blank prompt)
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_0_0.png
- text: a sks stuffed animal in the jungle
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_1_0.png
- text: a sks stuffed animal in the snow
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_2_0.png
- text: a sks stuffed animal on the beach
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_3_0.png
- text: a sks stuffed animal on a cobblestone street
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_4_0.png
- text: a sks stuffed animal on top of pink fabric
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_5_0.png
- text: a sks stuffed animal on top of a wooden floor
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_6_0.png
- text: a sks stuffed animal with a city in the background
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_7_0.png
- text: a sks stuffed animal with a mountain in the background
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_8_0.png
- text: a sks stuffed animal with a blue house in the background
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_9_0.png
- text: a sks stuffed animal on top of a purple rug in a forest
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_10_0.png
- text: a sks stuffed animal with a wheat field in the background
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_11_0.png
- text: a sks stuffed animal with a tree and autumn leaves in the background
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_12_0.png
- text: a sks stuffed animal with the Eiffel Tower in the background
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_13_0.png
- text: a sks stuffed animal floating on top of water
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_14_0.png
- text: a sks stuffed animal floating in an ocean of milk
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_15_0.png
- text: a sks stuffed animal on top of green grass with sunflowers around it
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_16_0.png
- text: a sks stuffed animal on top of a mirror
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_17_0.png
- text: a sks stuffed animal on top of the sidewalk in a crowded street
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_18_0.png
- text: a sks stuffed animal on top of a dirt road
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_19_0.png
- text: a sks stuffed animal on top of a white rug
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_20_0.png
- text: a red sks stuffed animal
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_21_0.png
- text: a purple sks stuffed animal
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_22_0.png
- text: a shiny sks stuffed animal
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_23_0.png
- text: a wet sks stuffed animal
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_24_0.png
- text: a cube shaped sks stuffed animal
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_25_0.png
- text: a photo of a sks stuffed animal
parameters:
negative_prompt: blurry, cropped, ugly
output:
url: ./assets/image_26_0.png
simpletuner-lora
This is a LyCORIS adapter derived from black-forest-labs/FLUX.1-dev.
The main validation prompt used during training was:
a photo of a sks stuffed animal
Validation settings
- CFG:
3.0 - CFG Rescale:
0.0 - Steps:
20 - Sampler:
FlowMatchEulerDiscreteScheduler - Seed:
42 - Resolution:
1024x1024 - 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
- a sks stuffed animal in the jungle
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal in the snow
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal on the beach
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal on a cobblestone street
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal on top of pink fabric
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal on top of a wooden floor
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal with a city in the background
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal with a mountain in the background
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal with a blue house in the background
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal on top of a purple rug in a forest
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal with a wheat field in the background
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal with a tree and autumn leaves in the background
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal with the Eiffel Tower in the background
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal floating on top of water
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal floating in an ocean of milk
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal on top of green grass with sunflowers around it
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal on top of a mirror
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal on top of the sidewalk in a crowded street
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal on top of a dirt road
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a sks stuffed animal on top of a white rug
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a red sks stuffed animal
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a purple sks stuffed animal
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a shiny sks stuffed animal
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a wet sks stuffed animal
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a cube shaped sks stuffed animal
- Negative Prompt
- blurry, cropped, ugly
- Prompt
- a photo of a sks stuffed animal
- 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: 0
- Training steps: 500
- Learning rate: 0.0001
- Learning rate schedule: polynomial
- Warmup steps: 100
- Max grad value: 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'])
- Optimizer: adamw_bf16
- Trainable parameter precision: Pure BF16
- Base model precision:
no_change - Caption dropout probability: 10.0%
LyCORIS Config:
{
"algo": "lora",
"multiplier": 1.0,
"linear_dim": 64,
"linear_alpha": 32,
"apply_preset": {
"target_module": [
"Attention",
"FeedForward"
],
"module_algo_map": {
"Attention": {
"factor": 16
},
"FeedForward": {
"factor": 8
}
}
}
}
Datasets
dreambooth-subject
- Repeats: 1000
- Total number of images: 5
- Total number of aspect buckets: 1
- Resolution: 1.048576 megapixels
- Cropped: False
- Crop style: None
- Crop aspect: None
- Used for regularisation data: No
dreambooth-subject-512
- Repeats: 1000
- Total number of images: 5
- 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
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 = 'black-forest-labs/FLUX.1-dev'
adapter_repo_id = 'tz2026/simpletuner-lora'
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.transformer)
wrapper.merge_to()
prompt = "a photo of a sks stuffed animal"
## 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.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
model_output = 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=1024,
height=1024,
guidance_scale=3.0,
).images[0]
model_output.save("output.png", format="PNG")