quilt-1m-finetuned-sd3.5-v2

quilt-1m-finetuned-sd3.5-v2

This is a full rank finetune derived from sd3/unknown-model.

The main validation prompt used during training was:

A high-resolution photomicrograph of histopathology tissue showing cellular architecture

Validation settings

• CFG: 3.0
• CFG Rescale: 0.0
• Steps: 20
• Sampler: FlowMatchEulerDiscreteScheduler
• Seed: 42
• Resolution: 1024x1024
• Skip-layer guidance: skip_guidance_layers=[7, 8, 9],

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: 0
• Training steps: 7000
• Learning rate: 0.0001
  • Learning rate schedule: cosine
  • Warmup steps: 200
• Max grad value: 0.1
• Effective batch size: 16
  • Micro-batch size: 1
  • Gradient accumulation steps: 4
  • Number of GPUs: 4
• Gradient checkpointing: True
• Prediction type: flow_matching (extra parameters=['flow_schedule_auto_shift', 'shift=0.0'])
• Optimizer: adamw_bf16
• Trainable parameter precision: Pure BF16
• Base model precision: no_change
• Caption dropout probability: 0.05%

Datasets

images-512

• Repeats: 1
• Total number of images: ~417748
• Total number of aspect buckets: 1
• Resolution: 0.262144 megapixels
• Cropped: True
• Crop style: random
• Crop aspect: square
• Used for regularisation data: No

images-768

• Repeats: 1
• Total number of images: ~266740
• Total number of aspect buckets: 1
• Resolution: 0.589824 megapixels
• Cropped: True
• Crop style: random
• Crop aspect: square
• Used for regularisation data: No

images-1024

• Repeats: 1
• Total number of images: ~246816
• 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

model_id = 'Minh-Ha/quilt-1m-finetuned-sd3.5-v2'
pipeline = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.bfloat16) # loading directly in bf16

prompt = "A high-resolution photomicrograph of histopathology tissue showing cellular architecture"
negative_prompt = 'blurry, cropped, ugly'

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=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,
    skip_guidance_layers=[7, 8, 9],
).images[0]

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