Sign In

flux1 bitte guidance bnb nf4 (AIO) 4 steps, 8 steps

Updated: Mar 13, 2025

base model

Download

1 variant available

nf4

4-bit normalized • 10.7 GB

Verified:

Type

Checkpoint Trained

Stats

262

Reviews

Published

Mar 12, 2025

Base Model

Flux.1 D

Hash

AutoV2
FA0E0F68FD
First Birthday Badge

sabin

The FLUX.1 [dev] Model is licensed by Black Forest Labs. Inc. under the FLUX.1 [dev] Non-Commercial License. Copyright Black Forest Labs. Inc.

IN NO EVENT SHALL BLACK FOREST LABS, INC. BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH USE OF THIS MODEL.

Creator: silveroxides https://huggingface.co/silveroxides/flux1-nf4-weights

Flux model that can work with Flux.D loras at 4-8 steps. The model works with 20+ steps with no problem, and somehow works even with 2 steps.

This file model contains nipple censoring like a vanilla flux

AIO/AllInOne: VAE and text encoder are included.

I have been using this model on a regular basis for the past half a year. The author never posted it here, but I think it deserves attention. Unfortunately, even if you specify less than 20 steps in the online civitai generator it will not reduce the price in buzz.

Don't forget to set "Diffusion in Low Bits" to "Automatic (fp16 LoRA)" when you actually use LoRAs