import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("CabalResearch/NoobAI-V-pred-to-Flow-LoRA", dtype=torch.bfloat16, device_map="cuda")
prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k"
image = pipe(prompt).images[0]
V-pred to Flow LoRA
A LoRA-based conversion training, based on NoobAI V-pred 1.0. Allows v-pred models to utilize flow-based prediction.
You can find basic workflow alongside models.
Possible Benefits
- Reduction in color bias
- Ability to use flow-specific features and samplers
Possible Downsides
- Can desaturate too much
Caveats
- Results are model-dependent, your mileage may vary.
- Is only a LoRA adaptation, will not provide full Flow benefits.
Support
If you wish to support our continuous effort of making waifus 0.2% better, you can do it here:
- Downloads last month
- 18
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for CabalResearch/NoobAI-V-pred-to-Flow-LoRA
Base model
Laxhar/noobai-XL-1.0 Finetuned
Laxhar/noobai-XL-Vpred-0.5 Finetuned
Laxhar/noobai-XL-Vpred-0.6 Finetuned
Laxhar/noobai-XL-Vpred-0.65 Finetuned
Laxhar/noobai-XL-Vpred-0.75 Finetuned
Laxhar/noobai-XL-Vpred-1.0