This repository aims to add support for various random image diffusion models to ComfyUI.
Simply clone this repo to your custom_nodes folder using the following command: git clone https://github.com/city96/ComfyUI_ExtraModels custom_nodes/ComfyUI_ExtraModels.
I haven't tested this completely, so if you know what you're doing, use the regular venv/git clone install option when installing ComfyUI.
Go to the where you unpacked ComfyUI_windows_portable to (where your run_nvidia_gpu.bat file is) and open a command line window. Press CTRL+SHIFT+Right click in an empty space and click "Open PowerShell window here".
In case you haven't installed in through the manager, run git clone https://github.com/city96/ComfyUI_ExtraModels .\ComfyUI\custom_nodes\ComfyUI_ExtraModels
To install the requirements on windows, run these commands in the same window:
.\python_embeded\python.exe -m pip install -r .\ComfyUI\custom_nodes\ComfyUI_ExtraModels\requirements.txt
.\python_embeded\python.exe -m pip install bitsandbytes --prefer-binary --extra-index-url=https://jllllll.github.io/bitsandbytes-windows-webui
To update, open the command line window like before and run the following commands:
cd .\ComfyUI\custom_nodes\ComfyUI_ExtraModels\
git pull
Alternatively, use the manager, assuming it has an update function.
- Uses class labels instead of prompts
- Limited to 256x256 or 512x512 images
- Same latent space as SD1.5 (works with the SD1.5 VAE)
- Works in FP16, but no other optimization (yet)
- Download the original model weights from the DiT Repo or the converted FP16 safetensor ones from Huggingface.
- Place them in
ComfyUI\models\dit(created on first run after installing the extension) - Load the model and select the class labels as shown in the image below
- Make sure to use the Empty label conditioning for the Negative input of the KSampler!
ConditioningCombine nodes should work for combining multiple labels. The area ones don't since the model currently can't handle dynamic input dimensions.
- Uses T5 text encoder instead of clip
- Available in 512 and 1024 versions, needs specific pre-defined resolutions to work correctly
- Same latent space as SD1.5 (works with the SD1.5 VAE)
- Attention needs optimization, images look worse without xformers.
- Download the model weights from the PixArt alpha repo - you most likely want the 1024px one -
PixArt-XL-2-1024-MS.pth - Place them in your checkpoints folder
- Load them with the correct PixArt checkpoint loader
- Follow the T5 section of this readme to set up the T5 text encoder
Limitations:
- The default
KSampleruses a different noise schedule/sampling algo (I think), so it most likely won't work as expected. PixArt DPM Samplerrequires the negative prompt to be shorter than the positive prompt.PixArt DPM Samplercan only work with a batch size of 1.PixArt T5 Text Encodeis from the reference implementation, therefore it doesn't support weights.T5 Text Encodesupport weights, but I can't attest to the correctness of the implementation.
PixArt uses the same T5v1.1-xxl text encoder as DeepFloyd, so the T5 section of the readme also applies.
Important
Make sure to pip install timm==0.6.13 (or alternatively, pip install -r requirements.txt in the folder where the extension is)
Installing xformers is optional but strongly recommended as torch SDP is only partially implemented, if that.
The model files can be downloaded from the DeepFloyd/t5-v1_1-xxl repository.
You will need to download the following 4 files:
config.jsonpytorch_model-00001-of-00002.binpytorch_model-00002-of-00002.binpytorch_model.bin.index.json
Place them in your ComfyUI/models/t5 folder. You can put them in a subfolder called "t5-v1.1-xxl" though it doesn't matter. There are int8 safetensor files in the other DeepFloyd repo, thought they didn't work for me.
Loaded onto the CPU, it'll use about 22GBs of system RAM. Depending on which weights you use, it might use slightly more during loading.
Loaded in bnb4bit mode, it only takes around 6GB VRAM, making it work with 12GB cards. The only drawback is that it'll constantly stay in VRAM since BitsAndBytes doesn't allow moving the weights to the system RAM temporarily. Switching to a different workflow should still release the VRAM as expected. Pascal cards (1080ti, P40) seem to struggle with 4bit. Select "cpu" if you encounter issues.
On windows, you may need a newer version of bitsandbytes for 4bit. Try python -m pip install bitsandbytes --prefer-binary --extra-index-url=https://jllllll.github.io/bitsandbytes-windows-webui
Important
You may also need to upgrade transformers and install spiece for the tokenizer. pip install -r requirements.txt
A few custom VAE models are supported. The option to select a different dtype when loading is also possible, which can be useful for testing/comparisons.
This now works thanks to the work of @mrsteyk and @madebyollin - Gist with more info.
- Download the converted safetensor VAE from this HF repository. If you downloaded the OpenAI model before, it won't work, as it is a TorchScript file. Feel free to delete it.
- Put the file in your VAE folder
- Load it with the ExtraVAELoader
- Set it to fp16 or bf16 to not run out of VRAM
- Use tiled VAE decode if required
kl-f4/8/16/32 from the compvis/latent diffusion repo.
vq-f4/8/16 from the taming transformers repo, weights for both vq and kl models available here
vq-f8 can accepts latents from the SD unet but just like xl with v1 latents, output largely garbage. The rest are completely useless without a matching UNET that uses the correct channel count.
