import torch
from diffusers import DiffusionPipeline
from diffusers.utils import load_image
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("black-forest-labs/FLUX.2-klein-4B", dtype=torch.bfloat16, device_map="cuda")
pipe.load_lora_weights("xocialize/object-remove-FLUX.2-klein-4B-lora")
prompt = "Remove the highlighted object from the scene"
input_image = load_image("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/cat.png")
image = pipe(image=input_image, prompt=prompt).images[0]Flux Object Removal — FLUX.2-klein-4B LoRA (mirror)
A rank-16 edit LoRA for FLUX.2-klein-4B that removes a highlighted object from an image and fills the region with contextually appropriate content (inpainting-style removal).
This is a verbatim mirror, not original work. All credit belongs to the original author, fal (fal.ai). This repository exists so downstream code can depend on a stable, long-lived URL; the file is byte-for-byte identical to the source (verify the SHA-256 below).
Attribution & source
- Original author: fal (fal.ai)
- Source: https://huggingface.co/fal/flux-2-klein-4B-object-remove-lora
- Live demo (author): https://fal.ai/models/fal-ai/flux-2/klein/4b/base/edit/lora
- License: Apache-2.0, as published by fal on the source repository. Mirrored here
under those terms with attribution retained (see
NOTICE). If you are the author and want this mirror changed or removed, open a discussion and it will be actioned promptly.
Provenance (verify the mirror)
| File | flux-object-remove-lora.safetensors |
| Size | 58,441,728 bytes (~58 MB) |
| SHA-256 | 70CEEA34A8BCD6E7D0249BA26BB59A3CD8B8C38745FAB65A3EB112A5C13E3C69 |
| Tensors | 136 (f32) |
| Rank | 16 |
| Base model | black-forest-labs/FLUX.2-klein-4B |
| Key dialect | BFL-fused with PEFT base_model.model. prefix |
The source repo also ships a ComfyUI-converted variant
(..._comfy_converted.safetensors, diffusion_model. prefix); this mirror carries the
canonical diffusers/PEFT file. Both map to the same 40 double + 40 single (80) targets.
Usage
- Base: FLUX.2-klein-4B (image edit). Mark/highlight the object to remove in the input.
- Prompt:
Remove the highlighted object from the scene - Strength (scale): ~1.1 (per the author's fal.ai SDK example).
Disclaimer
Mirrored in good faith based on fal's Apache-2.0 release. No affiliation with or endorsement by fal is implied.
- Downloads last month
- -
Model tree for xocialize/object-remove-FLUX.2-klein-4B-lora
Base model
black-forest-labs/FLUX.2-klein-4B