Zuellni-ComfyUI-ExLlama-Nodes/README.md

ComfyUI ExLlama Nodes

A simple text generator for ComfyUI utilizing ExLlamaV2.

Installation

Navigate to the root ComfyUI directory, clone the repository to custom_nodes and install dependencies:

git clone https://github.com/Zuellni/ComfyUI-ExLlama-Nodes custom_nodes/ComfyUI-ExLlama-Nodes
pip install -r custom_nodes/ComfyUI-ExLlama-Nodes/requirements.txt

Optionally, you can install flash-attention by uncommenting the relevant lines in the requirements file. It should lower VRAM usage but your mileage may vary.

Important

The wheels included in the requirements file should match the latest portable ComfyUI build. If you see any ExLlama-related errors while loading the nodes, try to install it manually following the official instructions.

Usage

Only EXL2 and 4-bit GPTQ models are supported. You can find a lot of them on Hugging Face. Refer to the model card in each repository for details about quant differences and instruction formats.

To use a model with the nodes, you should clone its repository with git or manually download all the files and place them in models/llm. For example, if you'd like to download the 4-bit 32g version of Zephyr 7B Beta, use the following command:

git clone https://huggingface.co/TheBloke/zephyr-7B-beta-GPTQ -b gptq-4bit-32g-actorder_True models/llm/zephyr-7b-gptq-32g

Tip

You can add your own llm path to the extra_model_paths.yaml file and place the models there instead.

Nodes

Name	Description
Loader	Loads models from the llm directory. gpu_split - comma-separated VRAM in MB per GPU, if using more than one. cache_8bit - lower VRAM usage but also lower speed if set to True. max_seq_len - max context length, higher number equals higher VRAM usage. Setting it to 0 will make the model use default context length from its config file.
Generator	Generates text based on the given prompt. Refer to text-generation-webui for parameter explanations. unload - unloads the model after each generation if set to True, freeing all the VRAM used. single_line - stops generation on new line. max_tokens - max new tokens to generate, setting it to 0 will make the model use all available context.
Preview	Displays generated text in the UI.
Replace	Replaces variable names enclosed in brackets, such as [a], with their values.

Workflow

The example workflow is embedded in the image below and can be opened in ComfyUI.