pointerpointer-v1

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

The main validation prompt used during training was:

A bustling kitchen scene with three chefs in white uniforms and tall hats. The chef in the foreground, a woman with her hair in a tight bun, is pointing emphatically at a large pot on the stove with her right index finger. Her left hand holds a wooden spoon, and her expression is one of intense concentration. To her right, a male chef is chopping vegetables on a cutting board, but his head is turned towards where his colleague is pointing. In the background, the third chef is reaching up to a high shelf, pointing with his left hand at a specific spice jar. Steam rises from several pots on the industrial stove, and stainless steel appliances gleam in the bright kitchen lighting. The scene conveys a sense of urgency and coordination in a professional kitchen environment.

Validation settings

CFG: 3.0
CFG Rescale: 0.0
Steps: 25
Sampler: None
Seed: 42
Resolutions: 1024x1024,832x1216

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 bustling kitchen scene with three chefs in white uniforms and tall hats. The chef in the foreground, a woman with her hair in a tight bun, is pointing emphatically at a large pot on the stove with her right index finger. Her left hand holds a wooden spoon, and her expression is one of intense concentration. To her right, a male chef is chopping vegetables on a cutting board, but his head is turned towards where his colleague is pointing. In the background, the third chef is reaching up to a high shelf, pointing with his left hand at a specific spice jar. Steam rises from several pots on the industrial stove, and stainless steel appliances gleam in the bright kitchen lighting. The scene conveys a sense of urgency and coordination in a professional kitchen environment.

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: 3
Training steps: 100
Learning rate: 0.0002
Effective batch size: 90
- Micro-batch size: 45
- Gradient accumulation steps: 2
- Number of GPUs: 1
Prediction type: flow-matching
Rescaled betas zero SNR: False
Optimizer: adamw_bf16
Precision: bf16
Quantised: No
Xformers: Not used
LoRA Rank: 16
LoRA Alpha: None
LoRA Dropout: 0.1
LoRA initialisation style: default

Datasets

pointerpointer

Repeats: 0
Total number of images: 711
Total number of aspect buckets: 1
Resolution: 0.262144 megapixels
Cropped: True
Crop style: center
Crop aspect: square

Inference

import torch
from diffusers import DiffusionPipeline

model_id = 'black-forest-labs/FLUX.1-dev'
adapter_id = 'MaheshNat/pointerpointer-v1'
pipeline = DiffusionPipeline.from_pretrained(model_id)
pipeline.load_lora_weights(adapter_id)

prompt = "A bustling kitchen scene with three chefs in white uniforms and tall hats. The chef in the foreground, a woman with her hair in a tight bun, is pointing emphatically at a large pot on the stove with her right index finger. Her left hand holds a wooden spoon, and her expression is one of intense concentration. To her right, a male chef is chopping vegetables on a cutting board, but his head is turned towards where his colleague is pointing. In the background, the third chef is reaching up to a high shelf, pointing with his left hand at a specific spice jar. Steam rises from several pots on the industrial stove, and stainless steel appliances gleam in the bright kitchen lighting. The scene conveys a sense of urgency and coordination in a professional kitchen environment."

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=25,
    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=3.0,
).images[0]
image.save("output.png", format="PNG")