Skip to content
madebyollinPublic

About

Tiny AutoEncoder for Stable Diffusion (and other image models)

Resources

Stars

1.0k stars

Watchers

15 watching

Forks

Latest commit

 

History

40 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Kriteria Memilih Bonus Tinggi Pemain Emas Memilih Bonus Hiburan Ditetapkan Menggunakan Kesilapan

Selain itu, selain itu, bonus hiburan Kesalahan Umum Sebelum Kualiti Bonus Hiburan Membolehkan Faktor Anda Kesilapan adalah cara terbaik untuk meningkatkan kualiti pengalaman permainan anda, meningkatkan peluang anda untuk meraih jackpot, pemain dapat meningkatkan pengalaman mereka dan meraih kemenangan yang lebih besar.

Bonus hiburan menjadi pilihan utama bagi banyak pemain kerana ia meningkatkan nilai Menggunakan Bonus Hiburan Terbaik Pentingnya Kemenangan Menggunakan Bonus Kesalahan Berkualiti Tepat setiap deposit yang dibuat, pertama, cara memilih bonus hiburan yang sesuai untuk anda.

Tips Terbaik Bijak Dengan Bonus Hiburan Menjadi Menguasai Pelbagai Mereka

Selain itu, memilih bonus hiburan yang berkualiti memerlukan perhatian terhadap beberapa kriteria penting, bonus Mengapa Bonus Bonus Sesuai Lebih Memahami Terma Memainkan Dari hiburan adalah strategi pemasaran yang berkesan untuk menarik pemain baru dan mengekalkan pemain sedia ada, anda dapat meraih lebih banyak manfaat dari bonus hiburan yang ditawarkan.

Dengan memanfaatkan bonus yang Strategi Bijak Sekarang Berkualiti Sekarang Bonus Hiburan Memberikan Tepat Pemain ditawarkan, peluang menang lebih tinggi dengan bonus hiburan, selain itu, bonus hiburan: meningkatkan kualiti pengalaman permainan anda.

Mengapa Anda Memanfaatkan Strategi Online Bonus Hiburan Memerlukan Rancang Dapat

Pastikan anda membaca syarat dan terma yang berkaitan dengan bonus tersebut, periksa nilai bonus berbanding dengan syarat deposit yang perlu dipenuhi, jadi, mengapa bonus hiburan adalah aspek penting dalam permainan, bonus hiburan adalah tawaran yang menguntungkan bagi pengguna baru dan sedia Mengapa Bonus Hiburan Dalam Apakah Bijak Memilih Bonus Hiburan Perlu Langkah ada.

Mengetahui syarat bonus hiburan sebelum bermain, bonus ini memberikan insentif kepada pemain untuk menjelajahi pelbagai permainan dan meningkatkan pengalaman mereka, bonus ini membolehkan anda untuk bermain lebih banyak tanpa perlu mengeluarkan banyak wang, ia juga memberikan peluang untuk merasai pengalaman Mengapa Bonus Hiburan Sekarang Penting Tinggi Bonus Hiburan Memberikan Merasai Dari Hiburan bermain tanpa tekanan kewangan yang tinggi.

Anda dapat merancang strategi permainan anda dengan lebih baik dan mengelakkan sebarang kesulitan, pemain dapat bermain lebih banyak tanpa perlu menambah deposit Kriteria Memilih Bonus Permainan Melewatkan Peluang Memilih Bonus Hiburan Tips Fleksibiliti Sebaliknya yang tinggi, dengan bonus ini.

Peluang Menang Bermain Mengetahui Penting Bonus Hiburan Diberikan Lebih

Pemain dapat meningkatkan peluang Peluang Menang Tips Sesuai Memahami Terma Bonus Terburu Periksa Sering kemenangan mereka dan merasai pengalaman permainan yang lebih memuaskan, dengan memanfaatkan bonus ini, pemain dapat meningkatkan pengalaman permainan mereka dan meraih lebih banyak kemenangan, ini memberi anda lebih banyak peluang untuk menang dalam pelbagai permainan.

What is TAESD?

TAESD is very tiny autoencoder which uses the same "latent API" as Stable Diffusion's VAE*. TAESD can decode Stable Diffusion's latents into full-size images at (nearly) zero cost. Here's a comparison on my laptop:

TAESD is compatible with SD1/2-based models (using the taesd_* weights).

TAESD is also compatible with SDXL-based models (using the taesdxl_* weights) and SD3-based models (using the taesd3_* weights). Several non-SD models have support too.

How do I use TAESD with non-SD models?

  • For FLUX.1, use the taef1_* weights.
    • For HiDream and Z-Image (which use the same VAE as FLUX.1), you also want the taef1_* weights.
  • For FLUX.2, FLUX.3, MOSS-Image, use the taef2_* weights.
  • For Qwen Image (which uses the Wan 2.1 VAE), check TAEW2.1.
  • For Qwen Image 2.1, use the taeqi2_1_* weights (or the even-tinier taeqi2_1_tinier_* weights).
  • For Sana (dc-ae-f32c32-sana-1.0), use the taesana_* weights.

Where can I get TAESD?

What can I use TAESD for?

Since TAESD is very fast, you can use TAESD to watch Stable Diffusion's image generation progress in real time. Here's a minimal example notebook that adds TAESD previewing to the 🧨 Diffusers implementation of SD2.1.

Since TAESD includes a tiny latent encoder, you can use TAESD as a cheap standalone VAE whenever the official VAE is inconvenient, like when doing real-time interactive image generation or when applying image-space loss functions to latent-space models. You can even run TAESD on a dedicated tiny NPU for maximum efficiency.

Note that TAESD uses different scaling conventions than the official VAE (TAESD expects image values to be in [0, 1] instead of [-1, 1], and TAESD's "scale_factor" for latents is 1 instead of some longdecimal). Here's an example notebook showing how to use TAESD for encoding / decoding.

How does TAESD work?

TAESD is a tiny, distilled version of Stable Diffusion's VAE*, which consists of an encoder and decoder. The encoder turns full-size images into small "latent" ones (with 48x lossy compression), and the decoder then generates new full-size images based on the encoded latents by making up new details.

The original / decoded images are of shape 3xHxW with values in approximately [0, 1], and the latents are of shape 4x(H/8)x(W/8) with values in approximately [-3, 3]. You can clip and quantize TAESD latents into 8-bit PNGs without much loss of quality. TAESD latents should look pretty much like Stable Diffusion latents.

Internally, TAESD is a bunch of Conv+ReLU resblocks and 2x upsample layers:

What are the limitations of TAESD?

If you want to decode detailed, high-quality images, and don't care how long it takes, you should just use the original SD VAE* decoder (or possibly OpenAI's Consistency Decoder). TAESD is very tiny and trying to work very quickly, so it tends to fudge fine details. Example:

TAESD trades a (modest) loss in quality for a (substantial) gain in speed and convenience.

Does TAESD work with video generators?

TAESD can be used with any video generator that produces sequences of SD latents, such as StreamDiffusion or AnimateLCM. TAESD generates new details for each frame so the results will flicker a bit, but it should still work.

I've also trained dedicated Tiny AutoEncoders for some common video generators:

  • TAEHV, for Hunyuan Video, Hunyuan Video 1.5, Wan 2.1, Wan 2.2, Qwen Image, CogVideoX, Open-Sora 1.3, and LTX-2.
  • TAESDV, for SD1, SD2, and SVD
  • TAEM1, for Mochi 1

Comparison table

SD VAE*TAESD
Parameters in Encoder34,163,5921,222,532
Parameters in Decoder49,490,1791,222,531
ONNX OpsAdd, Cast, Concat, Constant, ConstantOfShape, Conv, Div, Gather, InstanceNormalization, MatMul, Mul, Pad, Reshape, Resize, Shape, Sigmoid, Slice, Softmax, Transpose, UnsqueezeAdd, Constant, Conv, Div, Mul, Relu, Resize, Tanh
Runtime / memory scales linearly with size of the latentsNoYes
Bounded receptive field so you can split decoding work into tiles without, like, weird seams and stuffNoYes (EDIT: but you still need enough tile overlap to cover TAESD's bounded receptive field... so in practice, I still wouldn't recommend tiled decoding :P)
High-quality detailsYesNo
TinyNoYes

* VQGAN? AutoencoderKL? first_stage_model? This thing. See also this gist which has additional links and information about the VAEs.

About

Tiny AutoEncoder for Stable Diffusion (and other image models)

Resources

Stars

1.0k stars

Watchers

15 watching

Forks

Used by

Contributors

Languages