Hugging Face's logo

bghira
/
flux-LoRA-r16-SOAP-4e-4-bs4

Text-to-Image
Diffusers
flux
flux-diffusers
simpletuner
safe-for-work
lora
template:sd-lora
standard
Model card Files Community
flux-LoRA-r16-SOAP-4e-4-bs4 / README.md
bghira's picture
bghira
Model card auto-generated by SimpleTuner
eb09abd verified
preview code
|
raw
history blame contribute delete
6.87 kB
metadata
license: other
base_model: black-forest-labs/FLUX.1-dev
tags:
  - flux
  - flux-diffusers
  - text-to-image
  - diffusers
  - simpletuner
  - safe-for-work
  - lora
  - template:sd-lora
  - standard
inference: true
widget:
  - text: unconditional (blank prompt)
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_0_0.png
  - text: Distinguished Gentleman
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_1_0.png
  - text: >-
      A highly detailed portrait of a distinguished man in his 50s with sharp
      features and piercing blue eyes, wearing a tailored suit, set in a classic
      study, photorealistic, 8k resolution, dramatic lighting, oil painting
      style
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_2_0.png
  - text: Graceful Dancer
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_3_0.png
  - text: >-
      An artistic depiction of a graceful woman dancing under the moonlight in a
      serene forest, wearing a flowing white dress, soft lighting, dreamy
      atmosphere, high resolution, intricate details, impressionist style
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_4_0.png
  - text: Tommy Chong Portrait
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_5_0.png
  - text: >-
      A realistic portrait of Tommy Chong smiling warmly, wearing his signature
      bandana and glasses, in a relaxed setting, high detail, photorealistic, 8k
      resolution, natural lighting
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_6_0.png
  - text: Cheech Marin Portrait
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_7_0.png
  - text: >-
      A high-definition portrait of Cheech Marin with a mischievous grin,
      wearing a colorful shirt, set against a lively urban backdrop,
      photorealistic, ultra-detailed, vibrant colors
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_8_0.png
  - text: Cheech and Chong Duo
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_9_0.png
  - text: Cheech Marin and Tommy Chong standing together, smiling, iconic pose
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_10_0.png
  - text: Cheech and Chong Journey
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_11_0.png
  - text: >-
      An intricate scene of Cheech Marin and Tommy Chong driving a vintage van
      along a coastal highway at sunset, laughing and enjoying the journey,
      highly detailed, cinematic lighting, 8k resolution, dynamic angle
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_12_0.png
  - text: Adventurous Explorer
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_13_0.png
  - text: >-
      An epic scene of a rugged man standing atop a snow-capped mountain peak at
      sunrise, windswept hair, determined gaze, holding a vintage compass,
      surrounded by breathtaking landscapes, ultra-realistic, high dynamic
      range, panoramic view
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_14_0.png
  - text: Ethereal Ballerina
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_15_0.png
  - text: >-
      A captivating image of a ballerina performing on a grand stage,
      illuminated by soft golden lights, wearing a delicate tutu, graceful pose,
      photorealistic, high resolution, dreamlike atmosphere
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_16_0.png
  - text: Comedy Duo Live
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_17_0.png
  - text: >-
      Cheech and Chong performing live on stage with microphones, energetic
      crowd, spotlight on them, vibrant colors, high-definition
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_18_0.png
  - text: Retro Duo Poster
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_19_0.png
  - text: >-
      A retro-style movie poster featuring Cheech Marin and Tommy Chong in a
      comedic pose, bold colors, vintage fonts, high detail, artistic
      illustration, 1970s aesthetic
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_20_0.png
  - text: tommy chong sitting to the left of cheech marin
    parameters:
      negative_prompt: blurry, cropped, ugly
    output:
      url: ./assets/image_21_0.png

flux-LoRA-r16-SOAP-4e-4-bs4

This is a standard PEFT LoRA derived from black-forest-labs/FLUX.1-dev.

The main validation prompt used during training was:

tommy chong sitting to the left of cheech marin

Validation settings

  • CFG: 3.0
  • CFG Rescale: 0.0
  • Steps: 20
  • Sampler: None
  • Seed: 42
  • Resolution: 512x512

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
Distinguished Gentleman
Negative Prompt
blurry, cropped, ugly
Prompt
A highly detailed portrait of a distinguished man in his 50s with sharp features and piercing blue eyes, wearing a tailored suit, set in a classic study, photorealistic, 8k resolution, dramatic lighting, oil painting style
Negative Prompt
blurry, cropped, ugly
Prompt
Graceful Dancer
Negative Prompt
blurry, cropped, ugly
Prompt
An artistic depiction of a graceful woman dancing under the moonlight in a serene forest, wearing a flowing white dress, soft lighting, dreamy atmosphere, high resolution, intricate details, impressionist style
Negative Prompt
blurry, cropped, ugly
Prompt
Tommy Chong Portrait
Negative Prompt
blurry, cropped, ugly
Prompt
A realistic portrait of Tommy Chong smiling warmly, wearing his signature bandana and glasses, in a relaxed setting, high detail, photorealistic, 8k resolution, natural lighting
Negative Prompt
blurry, cropped, ugly
Prompt
Cheech Marin Portrait
Negative Prompt
blurry, cropped, ugly
Prompt
A high-definition portrait of Cheech Marin with a mischievous grin, wearing a colorful shirt, set against a lively urban backdrop, photorealistic, ultra-detailed, vibrant colors
Negative Prompt
blurry, cropped, ugly
Prompt
Cheech and Chong Duo
Negative Prompt
blurry, cropped, ugly
Prompt
Cheech Marin and Tommy Chong standing together, smiling, iconic pose
Negative Prompt
blurry, cropped, ugly
Prompt
Cheech and Chong Journey
Negative Prompt
blurry, cropped, ugly
Prompt
An intricate scene of Cheech Marin and Tommy Chong driving a vintage van along a coastal highway at sunset, laughing and enjoying the journey, highly detailed, cinematic lighting, 8k resolution, dynamic angle
Negative Prompt
blurry, cropped, ugly
Prompt
Adventurous Explorer
Negative Prompt
blurry, cropped, ugly
Prompt
An epic scene of a rugged man standing atop a snow-capped mountain peak at sunrise, windswept hair, determined gaze, holding a vintage compass, surrounded by breathtaking landscapes, ultra-realistic, high dynamic range, panoramic view
Negative Prompt
blurry, cropped, ugly
Prompt
Ethereal Ballerina
Negative Prompt
blurry, cropped, ugly
Prompt
A captivating image of a ballerina performing on a grand stage, illuminated by soft golden lights, wearing a delicate tutu, graceful pose, photorealistic, high resolution, dreamlike atmosphere
Negative Prompt
blurry, cropped, ugly
Prompt
Comedy Duo Live
Negative Prompt
blurry, cropped, ugly
Prompt
Cheech and Chong performing live on stage with microphones, energetic crowd, spotlight on them, vibrant colors, high-definition
Negative Prompt
blurry, cropped, ugly
Prompt
Retro Duo Poster
Negative Prompt
blurry, cropped, ugly
Prompt
A retro-style movie poster featuring Cheech Marin and Tommy Chong in a comedic pose, bold colors, vintage fonts, high detail, artistic illustration, 1970s aesthetic
Negative Prompt
blurry, cropped, ugly
Prompt
tommy chong sitting to the left of cheech marin
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: 12
  • Training steps: 1000
  • Learning rate: 0.0008
  • Effective batch size: 4
    • Micro-batch size: 4
    • Gradient accumulation steps: 1
    • Number of GPUs: 1
  • Prediction type: flow-matching
  • Rescaled betas zero SNR: False
  • Optimizer: soap
  • Precision: Pure BF16
  • Quantised: Yes: int8-quanto
  • Xformers: Not used
  • LoRA Rank: 16
  • LoRA Alpha: None
  • LoRA Dropout: 0.1
  • LoRA initialisation style: default

Datasets

cheechandchong-512

  • Repeats: 10
  • Total number of images: 17
  • Total number of aspect buckets: 3
  • Resolution: 0.262144 megapixels
  • Cropped: True
  • Crop style: random
  • Crop aspect: square

Inference 

import torch
from diffusers import DiffusionPipeline

model_id = 'black-forest-labs/FLUX.1-dev'
adapter_id = 'bghira/flux-LoRA-r16-SOAP-4e-4-bs4'
pipeline = DiffusionPipeline.from_pretrained(model_id)
pipeline.load_lora_weights(adapter_id)

prompt = "tommy chong sitting to the left of cheech marin"

pipeline.to('cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu')
image = 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(1641421826),
    width=512,
    height=512,
    guidance_scale=3.0,
).images[0]
image.save("output.png", format="PNG")