Instructions to use ionden/taew2.1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use ionden/taew2.1 with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir taew2.1 ionden/taew2.1
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Add provenance README
Browse files
README.md
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
library_name: mlx
|
| 4 |
+
tags:
|
| 5 |
+
- tiny-autoencoder
|
| 6 |
+
- taehv
|
| 7 |
+
- taew2.1
|
| 8 |
+
- qwen-image
|
| 9 |
+
- wan2.1
|
| 10 |
+
- mlx
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
# taew2.1 (re-host)
|
| 14 |
+
|
| 15 |
+
A faithful re-host of madebyollin's **taew2.1** — the Tiny AutoEncoder for the Wan 2.1 VAE
|
| 16 |
+
(used by Qwen-Image / Qwen-Image-Edit) — for use by
|
| 17 |
+
[`mlx-taef`](https://github.com/IonDen/mlx-taef) as a fast, low-memory preview decoder.
|
| 18 |
+
|
| 19 |
+
## Provenance
|
| 20 |
+
|
| 21 |
+
- **Canonical source:** [`madebyollin/taehv`](https://github.com/madebyollin/taehv) @ commit
|
| 22 |
+
`a1c8e6a2ba77b91f284ef98935ec5bd21a41d786`, file `safetensors/taew2_1.safetensors`.
|
| 23 |
+
- **sha256:** `04766eac0221b5390b985ae3fdcca652cbb4b1e8b82b28ea7ff89dfad1b1a93f` — byte-identical to
|
| 24 |
+
the canonical GitHub file. fp16, 128 tensors, one combined file holding both encoder and decoder.
|
| 25 |
+
- **License:** MIT (unchanged from upstream).
|
| 26 |
+
|
| 27 |
+
This repo exists only because the canonical weights are published on GitHub (madebyollin hosts no
|
| 28 |
+
Hugging Face copy), and re-uploaded community mirrors were found to differ from the canonical
|
| 29 |
+
weights. `mlx-taef` pins this repo by revision + sha256 so the bytes are verified at download time.
|