# ComfyUI Ollama

Custom [ComfyUI](https://github.com/comfyanonymous/ComfyUI) Nodes for interacting with [Ollama](https://ollama.com/) using the [ollama python client](https://github.com/ollama/ollama-python).

Integrate the power of LLMs into ComfyUI workflows easily or just experiment with LLM inference.

To use this properly, you would need a running Ollama server reachable from the host that is running ComfyUI.

<a href="https://www.buymeacoffee.com/stavsapq" target="_blank"><img src="https://cdn.buymeacoffee.com/buttons/default-orange.png" alt="Buy Me A Coffee" height="40" width="174"></a>

## Installation

Install ollama server on the desired host

<a href="https://ollama.com/" target="_blank">
  <img src="https://img.shields.io/badge/v0.4.2-green.svg?style=for-the-badge&labelColor=gray&label=Ollama&color=blue" alt=""/>
</a>

<a href="https://ollama.com/download/Ollama-darwin.zip" target="_blank">Download for macOS</a>

<a href="https://ollama.com/download/OllamaSetup.exe" target="_blank">Download for Windows</a>

Install on Linux

```shell 
curl -fsSL https://ollama.com/install.sh | sh
```

<a href="https://hub.docker.com/r/ollama/ollama" target="_blank">Docker Installation</a>

CPU only

```shell
docker run -d -p 11434:11434 -v ollama:/root/.ollama --name ollama ollama/ollama
```

NVIDIA GPU

```shell
docker run -d -p 11434:11434 --gpus=all -v ollama:/root/.ollama --name ollama  ollama/ollama
```

Use the [compfyui manager](https://github.com/ltdrdata/ComfyUI-Manager) "Custom Node Manager":

![pic](.meta/InstallViaManager.png)

Search `ollama` and select the one by `stavsap`

![pic](.meta/manager-install.png)

**Or**

1. git clone into the ```custom_nodes``` folder inside your ComfyUI installation or download as zip and unzip the contents to ```custom_nodes/compfyui-ollama```.
2. `pip install -r requirements.txt`
3. Start/restart ComfyUI

## Nodes

## V2 Release

![pic](.meta/v2.png)

### OllamaGenerateV2

A node that provides ability to set the system prompt and the prompt.

Ability to save context locally in the node `enable/disable`

Inputs:

- OllamaConnectivityV2 (optional)
- OllamaOptionsV2 (optional)
- images (optional)
- context (optional), a context from other OllamaConnectivityV2
- meta (optional), passing metadata of the OllamaConnectivityV2 and OllamaOptionsV2 from other OllamaGenerateV2 node.

### OllamaConnectivityV2

A node responsible only fot the connectivity to the ollama server

### OllamaOptionsV2

A node for full control of the ollama api options.

For an option to take effect, each option have also `enable/disable`, enabled options are passed to api call to ollama server.

Ollama API options can be found in [this table](https://github.com/ollama/ollama/blob/main/docs/modelfile.md#valid-parameters-and-values).

**Note**: There is an additional option `debug` that enables debug print in the cli, its not part of ollama api.

## V1 Release

### OllamaVision

A node that gives an ability to query input images. 

![pic](.meta/OllamaVision.png)

A model name should be model with Vision abilities, for example: https://ollama.com/library/llava.

### OllamaGenerate

A node that gives an ability to query an LLM via given prompt. 

![pic](.meta/OllamaGenerate.png)

### OllamaGenerateAdvance

A node that gives an ability to query an LLM via given prompt with fine tune parameters and an ability to preserve context for generate chaining. 

Check [ollama api docs](https://github.com/ollama/ollama/blob/main/docs/api.md#generate-a-completion) to get info on the parameters.

More [params info](https://github.com/ollama/ollama/blob/main/docs/modelfile.md#parameter)

![pic](.meta/generate-advance.png)

## Usage Example

Consider the following workflow of vision an image, and perform additional text processing with desired LLM. In the OllamaGenerate node set the prompt as input.

![pic](.meta/CombinedUsage1.png)

The custom Text Nodes in the examples can be found here: https://github.com/pythongosssss/ComfyUI-Custom-Scripts