Text-to-Video
Diffusers
Safetensors
modular_diffusers
vae
ltx2.3
lightricks
video-to-video
text-to-audio
Instructions to use AINovice2005/pruna-vaed-modular-diffusers with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use AINovice2005/pruna-vaed-modular-diffusers with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("AINovice2005/pruna-vaed-modular-diffusers", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
Update README.md
Browse files
README.md
CHANGED
|
@@ -5,4 +5,66 @@ library_name: diffusers
|
|
| 5 |
tags:
|
| 6 |
- modular_diffusers
|
| 7 |
- vae
|
| 8 |
-
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
tags:
|
| 6 |
- modular_diffusers
|
| 7 |
- vae
|
| 8 |
+
- ltx2.3
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
# Pruna VAE for Modular Diffusers
|
| 14 |
+
|
| 15 |
+
This repository provides a **Modular Diffusers** implementation of the [PrunaVAED](https://huggingface.co/PrunaAI/PrunaVAED). It packages the pruned decoder implementation together with the custom loading logic required to use
|
| 16 |
+
the model as a reusable Modular Diffusers component.
|
| 17 |
+
|
| 18 |
+
The repository exposes a custom `LoadPrunaVAE` block that downloads and instantiates `PrunaAutoencoderKLLTX2Video`, making it available as `components.vae` within a Modular Diffusers pipeline.
|
| 19 |
+
|
| 20 |
+
## Features
|
| 21 |
+
|
| 22 |
+
* Modular Diffusers block for `diffusers/LTX-2.3-Distilled-Diffusers`
|
| 23 |
+
* Custom `PrunaAutoencoderKLLTX2Video` implementation
|
| 24 |
+
* Compatible with `trust_remote_code=True`
|
| 25 |
+
* Self-contained repository containing both the implementation and model weights
|
| 26 |
+
* Intended for composition inside Modular Diffusers pipelines
|
| 27 |
+
|
| 28 |
+
## Intended Use
|
| 29 |
+
|
| 30 |
+
This repository is designed to be consumed by Modular Diffusers pipelines and reusable pipeline blocks. It is particularly useful for:
|
| 31 |
+
|
| 32 |
+
* Video inference
|
| 33 |
+
* Modular Diffusers experimentation
|
| 34 |
+
* Custom video generation pipelines
|
| 35 |
+
* Research on pruned video VAEs
|
| 36 |
+
|
| 37 |
+
## Usage
|
| 38 |
+
|
| 39 |
+
```python
|
| 40 |
+
from diffusers.modular_pipelines import ModularPipelineBlocks
|
| 41 |
+
|
| 42 |
+
blocks = ModularPipelineBlocks.from_pretrained(
|
| 43 |
+
"AINovice2005/pruna-vaed-modular-diffusers",
|
| 44 |
+
trust_remote_code=True,
|
| 45 |
+
)
|
| 46 |
+
|
| 47 |
+
pipeline = blocks.init_pipeline()
|
| 48 |
+
pipeline.load_components()
|
| 49 |
+
|
| 50 |
+
vae = pipeline.vae
|
| 51 |
+
```
|
| 52 |
+
|
| 53 |
+
The loaded component is an instance of:
|
| 54 |
+
|
| 55 |
+
```python
|
| 56 |
+
PrunaAutoencoderKLLTX2Video
|
| 57 |
+
```
|
| 58 |
+
|
| 59 |
+
and can be used anywhere a compatible LTX-2 Video VAE is expected.
|
| 60 |
+
|
| 61 |
+
|
| 62 |
+
> [!WARNING]
|
| 63 |
+
> The VAE decoder reconstructs frames from compressed latent representations.
|
| 64 |
+
> Depending on the downstream workflow, applying a dedicated image or video upscaler after decoding can noticeably improve perceived sharpness and fine detail, particularly for outputs intended for display at higher resolutions.
|
| 65 |
+
|
| 66 |
+
|
| 67 |
+
|
| 68 |
+
## Acknowledgements
|
| 69 |
+
|
| 70 |
+
This work builds upon the Pruna VAE implementation released by **PrunaAI** and adapts it for use as a reusable Modular Diffusers component.
|