---
license: other
license_name: qwen-research
license_link: https://huggingface.co/Qwen/Qwen-Image-2.1/blob/main/LICENSE
base_model: Qwen/Qwen-Image-2.1
library_name: diffusers
tags:
- vae
- qwen-image
---

# Texture-Fix-VAE-for-Qwen-Image-2.1

<em>Built with Qwen<sup>*</sup> (an unofficial finetune of the Qwen-Image-2.1 VAE)</em>

Texture-Fix-VAE-for-Qwen-Image-2.1 is the [Qwen-Image-2.1 VAE](https://huggingface.co/Qwen/Qwen-Image-2.1/tree/main/vae), but finetuned to produce cleaner textures with no checkerboard artifacts.

<details>
<summary>Comparison Settings</summary>

Texture-Fix-VAE-for-Qwen-Image-2.1's improved decoding is most noticeable in detailed, photo-style images.
The latents for the VAE comparison image below were generated by Qwen-Image-2.1 from the photo-style prompt:

> Landscape photograph of a subalpine wildflower meadow in the Pacific Northwest in midsummer: a clear mountain stream winding over mossy boulders through purple lupine and red paintbrush, dense old-growth Douglas fir and western red cedar forest behind, a snow-capped volcano in the distance, golden late-afternoon light, highly detailed

</details>

| Qwen-Image-2.1-VAE ([full-res](./images/decode-qwen-image-2.1-vae-full.png)) | 🪄 Texture-Fix-VAE-for-Qwen-Image-2.1 🪄 ([full-res](./images/decode-texture-fix-vae-full.png)) |
| --- | --- |
| ![](./images/decode-qwen-image-2.1-vae-zoomed-1.png) | ![](./images/decode-texture-fix-vae-zoomed-1.png) |
| ![](./images/decode-qwen-image-2.1-vae-zoomed-2.png) | ![](./images/decode-texture-fix-vae-zoomed-2.png) |
| ![](./images/decode-qwen-image-2.1-vae.png) | ![](./images/decode-texture-fix-vae.png) |

# Usage

**ComfyUI**

Download [`texture_fix_vae_for_qwen_image_2.1_bf16.safetensors`](./texture_fix_vae_for_qwen_image_2.1_bf16.safetensors) into `ComfyUI/models/vae/` and select it in the `Load VAE` node, in place of `qwen_image_2.1_vae_bf16.safetensors`.

**🧨 Diffusers**

```python
import torch
from diffusers import QwenImage21Pipeline, AutoencoderKLQwenImage21

vae = AutoencoderKLQwenImage21.from_pretrained("madebyollin/texture-fix-vae-for-qwen-image-2.1", torch_dtype=torch.bfloat16)
pipe = QwenImage21Pipeline.from_pretrained("Qwen/Qwen-Image-2.1", vae=vae, torch_dtype=torch.bfloat16).to("cuda")
```

# Mechanism

Texture-Fix-VAE-for-Qwen-Image-2.1 was created by finetuning the Qwen-Image-2.1 VAE decoder for ~5000 steps at learning rate 3e-5, with only the two highest-resolution decoder stages and output head unfrozen (7.5M trainable parameters), using the recipe developed for [TAESD](https://github.com/madebyollin/taesd).

The TAESD recipe, like most image autoencoder training recipes, uses a mix of PSNR-focused (MSE/MAE), LPIPS, and adversarial (GAN) loss terms.
Whenever precise details can't be reconstructed, MSE/MAE loss encourages blurring, LPIPS loss encourages blurring+checkerboarding (among other artifacts), and adversarial loss encourages generating sharp/plausible (but fake) detail without obvious artifacts.
This figure from [DC-AE](https://arxiv.org/abs/2410.10733) shows the importance of including adversarial (GAN) loss:

![Demo of the effects of adversarial loss, courtesy of the DC-AE paper](https://cdn-uploads.huggingface.co/production/uploads/630447d40547362a22a969a2/uj9vJSR34LyZEkQeL-VOU.png)

<sup>Figure: from *Deep Compression Autoencoder for Efficient High-Resolution Diffusion Models* (Chen et al., 2024, [arXiv:2410.10733](https://arxiv.org/abs/2410.10733)), licensed under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/); cropped to the first two rows.</sup>

I suspect the original Qwen-Image-2.1-VAE was trained without a working adversarial loss term.

# Metrics

Texture-Fix-VAE-for-Qwen-Image-2.1 makes perceptual quality metrics (rFID) better and reconstruction accuracy metrics (LPIPS/PSNR) slightly worse.

| Metric | Qwen-Image-2.1-VAE | Texture-Fix-VAE-for-Qwen-Image-2.1 |
| --- | --- | --- |
| rFID ↓ (COCO val2017, 5000 images @ 256²) | 3.37 | **2.08** |
| PSNR ↑ (COCO val2017 @ 256²) | **33.30** | 32.86 |
| LPIPS ↓ (COCO val2017 @ 256²) | **0.0357** | 0.0373 |
| PSNR ↑ (DIV2K valid, native 1024² crops) | **32.86** | 32.46 |
| LPIPS ↓ (DIV2K valid, native 1024² crops) | **0.0460** | 0.0480 |

# Attribution Notice

This fine-tuned VAE is based on Qwen/Qwen-Image-2.1; original materials © 2026 Hangzhou Tongyi Laboratory Technology Co., Ltd., licensed under the Qwen RESEARCH LICENSE AGREEMENT (see [LICENSE](./LICENSE)) for non-commercial/research use only. Built with Qwen<sup>*</sup>.

<sup>* In the sense that the **initial VAE weights** are from Qwen-Image. The decoder fine-tuning work was performed by `madebyollin` and Claude Opus.</sup>