import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("AcademiaSD/Boogu-Image-Edit-NF4-for-LoRA-Training", dtype=torch.bfloat16, device_map="cuda")
prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k"
image = pipe(prompt).images[0]Boogu Image Edit 0.1 — NF4 for LoRA training
Pre-quantized weights used by the Boogu Image Edit trainer of AcademiaSD LoRAlab Trainer Studio. The trainer downloads this repository on the first Pre-Cache; you don't need to download anything by hand.
| Folder | Content |
|---|---|
transformer/ |
Boogu Image Edit transformer, NF4 (bitsandbytes, ~5 GB) |
mllm/ |
Qwen3-VL-8B text encoder, NF4 (~5 GB, used only by Pre-Cache) |
vae/, processor/, scheduler/ |
Unchanged from the original repository |
boogu_nf4_metadata.json |
Hashes of the original weights, used to load the NF4 files with DiffSynth-Studio |
The NF4 files were made with DiffSynth-Studio's bitsandbytes_nf4 quantization and are meant for training; for inference use the original model.
Credits and license
- Boogu/Boogu-Image-0.1-Edit — Apache 2.0.
- DiffSynth-Studio by ModelScope — Apache 2.0.
NF4 conversion by AcademiaSD. Same license as the original model (Apache 2.0).
- Downloads last month
- -
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for AcademiaSD/Boogu-Image-Edit-NF4-for-LoRA-Training
Base model
Boogu/Boogu-Image-0.1-Edit