Hugging Face's logo

black-forest-labs
/
FLUX.2-small-decoder

Image-to-Image
Diffusers
Safetensors
Diffusion Single File
English
text-to-image
image-editing
flux
Model card Files
xet
 Community
2
huemin commited on
Commit
453b223
·
verified ·
1 Parent(s): 304d054

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -0
README.md CHANGED
@@ -14,6 +14,9 @@ tags:
14
 
15
  This repository holds a small (distilled) VAE decoder for FLUX.2. It is a drop-in replacement for the full decoder, with narrower channel widths (`[96, 192, 384, 384]` vs `[128, 256, 512, 512]`) for faster decoding with minimal quality loss. The encoder is unchanged.
16
 
 
 
 
17
  Compatible with all open-source FLUX.2 models:
18
  - [FLUX.2-klein-4B](https://huggingface.co/black-forest-labs/FLUX.2-klein-4B)
19
  - [FLUX.2-klein-9B](https://huggingface.co/black-forest-labs/FLUX.2-klein-9B)
 
14
 
15
  This repository holds a small (distilled) VAE decoder for FLUX.2. It is a drop-in replacement for the full decoder, with narrower channel widths (`[96, 192, 384, 384]` vs `[128, 256, 512, 512]`) for faster decoding with minimal quality loss. The encoder is unchanged.
16
 
17
+ - **~28M** decoder params (vs ~50M full decoder)
18
+ - **~1.4x** faster decode
19
+
20
  Compatible with all open-source FLUX.2 models:
21
  - [FLUX.2-klein-4B](https://huggingface.co/black-forest-labs/FLUX.2-klein-4B)
22
  - [FLUX.2-klein-9B](https://huggingface.co/black-forest-labs/FLUX.2-klein-9B)