Extensions/ComfyUI Custom Nodes for Latent Diffusion (CompVis F8 Large)
ComfyUI Extension

ComfyUI Custom Nodes for Latent Diffusion (CompVis F8 Large)

Native custom nodes to initialize and run original Latent Diffusion Model (LDM) F8 Large architecture from CompVis with a BERT text encoder.

By Winlensky·Created 22 days ago·Updated 16 days ago· 1
Winlensky/comfyui_compvis_ldm
Nodes7
On cloudLocal install
CategoryLDM
Stars1
Updated16 days ago

Nodes (7)

📜 LDM BERT Text Encode

Your prompt goes through BERT here, not CLIP — write sentences, not tag soup

LDM
🏰 Load LDM Checkpoint

The loader that resurrects the model Stable Diffusion came from

LDM
🌫️ LDM Empty Latent

This model was born at 256×256

LDM
🎲 LDM Sampler

There's no CFG here — meet the sampler that denoises the original LDM

LDM
💾 Save LDM Checkpoint

Shrink that 6 GB checkpoint to 3 GB — and snapshot the pipeline you actually run

LDM
🖼️ LDM VAE Decode

The last node in the chain, and the one with the magic number on it

LDM
📥 LDM VAE Encode

Pushing a real image into this model's latent space

LDM
Readme

ComfyUI Custom Nodes for Latent Diffusion (CompVis F8 Large)

<img width="1583" height="852" alt="image" src="https://github.com/user-attachments/assets/b76c6c5e-e0f9-49ca-a78a-93b0e882f990" />

A set of custom nodes for ComfyUI designed to natively initialize and run the historic original Latent Diffusion Model (LDM) F8 Large architecture from CompVis.

Unlike newer models (like Stable Diffusion 1.5), this specific pipeline relies on a heavy BERT-based text encoder (dim: 1280, vocab_size: 30522) and an aggressive f8 latent autoencoder. This repository enables standalone support for this unique model without relying on external .yaml configuration blueprints or old, deprecated codebases.

🚀 Key Features

  • Native LDM F8 Large Support: Full execution pipeline for text-to-image (txt2img) generation using the original architecture weights.
  • Lightweight Dependencies: No bloatware or heavy framework wrappers. Built on top of core packages (torch, safetensors, transformers) already shipped with standard ComfyUI setup.
  • Zero-Config Loading: The architecture is handled directly within the custom nodes backend, preventing configuration errors common with standard checkers.

📦 Installation

Support install in ComfyUI Manager or manual install for Windows and Linux

  1. Navigate to your ComfyUI custom nodes directory inside your terminal or PowerShell.
  2. Clone this repository: 
    git clone https://github.com/Winlensky/comfyui_compvis_ldm.git
  3. Install requirements (Most likely not needed): 
    cd comfyui_compvis_ldm
pip install -r requirements.txt
  4. Restart ComfyUI. The nodes will be registered automatically using your embedded Python environment.

🗂️ Model Downloads

  • Download Checkpoints: Latent Diffusion F8 Large on Hugging Face
  • Available Flavors:
    • latent-diffusion-f8-large-fp32.safetensors (Original precision, ~6GB)
    • latent-diffusion-f8-large-bf16.safetensors (Optimized precision, ~3GB)
    • latent-diffusion-f8-large-fp16.safetensors (Optimized precision, ~3GB)
    • latent-diffusion-f8-large-jack000-finetuned-fp16.safetensors (Finetuned version by jack000 for better structural composition)

📂 Where to place the files:

Place the downloaded .safetensors files strictly into your ComfyUI models directory under the dedicated LDM folder: ComfyUI\models\ldm

🛠️ Usage & Node Workflows

  • Resolution Notice: The LDM f8-large model was explicitly trained on downscaled, tight matrix resolutions. Ensure your target generation canvas is constrained (e.g., 256x256).

📜 Credits & License

  • Core architecture by CompVis (LMU Munich).
  • Finetuned variations by jack000.
  • This project is distributed under the MIT License.