Downloads · 30 days
48
9% of all-time downloads
Felldude/SDXL_HDR_VAE
SDXL_HDR_VAE is a machine learning model from Felldude. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for diffusers. The card lists the license as bigscience-openrail-m.
NOTE Encoder was not trained this VAE will likely not effect training in anyway
Downloads · 30 days
48
9% of all-time downloads
All-time downloads
529
Public
Parameters
83.7M
1 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors335 MB · 100%
From the Hugging Face model README
NOTE Encoder was not trained this VAE will likely not effect training in anyway
This repository contains evaluation results for an HDR VAE (High Dynamic Range Variational Autoencoder) designed for SDXL image reconstruction workflows.
The HDR VAE is evaluated against the BASE SDXL VAE using a multi-domain reconstruction analysis covering:
The objective is not only to measure pixel reconstruction accuracy, but to evaluate how well the VAE preserves the complete visual signal of an image through the encode/decode pipeline.
Measured through calibrated marker regions:
These evaluate whether reconstruction maintains luminance and color relationships.
| Metric | BASE SDXL VAE | HDR VAE |
|---|---|---|
| LPIPS ↓ | 0.0321 | 0.0593 |
| Gradient Energy Ratio | 0.792 | 1.167 |
| Color Support Ratio | 0.951 | 0.968 |
| Brightness Bias ↓ | 0.00101 | 0.00030 |
| Contrast Error ↓ | 0.00928 | 0.00439 |
| RGB Drift ↓ | 0.1608 | 0.1035 |
The BASE SDXL VAE achieves lower LPIPS error, indicating stronger agreement in learned perceptual feature space. Its reconstruction behavior favors smooth perceptual similarity.
The HDR VAE preserves more measurable image information:
The HDR VAE behaves as a more information-preserving reconstruction operator, maintaining image characteristics that are often reduced during standard VAE compression.
The HDR VAE is designed to improve reconstruction fidelity in areas where standard VAEs commonly lose information:
Rather than optimizing only for perceptual closeness, the HDR VAE emphasizes preservation of the original image signal through the latent representation.
The evaluation demonstrates two different reconstruction profiles:
BASE SDXL VAE
HDR VAE
The HDR VAE provides a reconstruction profile optimized for users requiring higher information retention, improved detail preservation, and more stable image characteristics through SDXL latent encoding and decoding.