https://github.com/h1ddenpr0cess20/ollamarama
Terminal-based AI chatbot with infinite personalities, using local LLMs via ollama
https://github.com/h1ddenpr0cess20/ollamarama
ai chatbot llama local-llm ollama terminal
Last synced: 11 months ago
JSON representation
Terminal-based AI chatbot with infinite personalities, using local LLMs via ollama
- Host: GitHub
- URL: https://github.com/h1ddenpr0cess20/ollamarama
- Owner: h1ddenpr0cess20
- License: agpl-3.0
- Created: 2023-12-22T00:49:26.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2025-03-12T20:16:26.000Z (over 1 year ago)
- Last Synced: 2025-03-12T21:24:41.563Z (over 1 year ago)
- Topics: ai, chatbot, llama, local-llm, ollama, terminal
- Language: Python
- Homepage:
- Size: 51.8 KB
- Stars: 7
- Watchers: 1
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Ollamarama
A terminal-based AI chatbot with infinite personalities, powered by local LLMs through Ollama. Create, customize, and chat with AI personalities directly from your terminal.
Also available for:
- [IRC](https://github.com/h1ddenpr0cess20/ollamarama-irc)
- [Matrix](https://github.com/h1ddenpr0cess20/ollamarama-matrix)
## Table of Contents
- [Features](#features)
- [Prerequisites](#prerequisites)
- [Installation](#installation)
- [Configuration](#configuration)
- [Usage](#usage)
- [Commands](#commands)
- [License](#license)
## Features
- Chat with locally-hosted LLMs using Ollama
- Create unlimited custom AI personalities
- Adjust model parameters like temperature and top_p on the fly
- Switch between different AI models
- Multi-line input support (Esc+Enter)
- Rich markdown rendering for AI responses
- Customizable system prompts
## Prerequisites
- Python 3.7 or higher
- [Ollama](https://ollama.com/) installed and running
- At least one LLM model pulled via Ollama
## Installation
1. Clone the repository:
```bash
git clone https://github.com/h1ddenpr0cess20/ollamarama.git
cd ollamarama
```
2. Install [Ollama](https://ollama.com/) if you haven't already:
```bash
curl https://ollama.com/install.sh | sh
```
3. Pull at least one model using Ollama:
```bash
ollama pull qwen2.5:14b # or any other model you prefer
```
4. Install Python dependencies:
```bash
pip install -r requirements.txt
```
## Configuration
Edit the `config.json` file to customize your setup:
## Usage
Run the application:
```bash
python3 ollamarama.py
```
Start chatting with the AI, or use commands to customize the experience.
## Commands
**Commands**
* `/help`: Shows the help menu
* `/reset`: Resets to default personality
* `/clear`: Resets and clears the screen
* `/stock`: Sets bot to stock model settings
* `/persona`: Activates personality changer (prompts for new personality)
* `/custom`: Use a custom system prompt
* `/model`: List models and change the current model
* `/model reset`: Reset to default model
* `/temperature`: Changes temperature setting
* `/top_p`: Changes top_p setting
* `/repeat_penalty`: Changes repeat_penalty setting
* `/quit` or `/exit`: Exits the program
**Tip:** Use Esc+Enter to input multiple lines of text.