🗨️ Ollama AI Chat Interface

Table of Contents

1. 📄 Description
2. 🔧 Prerequisites
3. 💻 Installation
4. ⚙️ Configuration
5. 🚀 Using the Interface
6. 📂 Managing Categories
7. 📝 History
8. ❓ Why JSON?
9. 🌐 Obtaining IPv4
10. 💡 Suggestions for Improvement
11. 👨‍💻 Author

📄 Description

This project is a web interface for chat between a user and an AI, utilizing various AI models compatible with the Ollama API. Supported models include: llama3, mistral, mini orca, and any LLMs using the same API.

🔧 Prerequisites

• Install Ollama and launch a compatible model.
• Supported models: llama3, mistral, mini orca, etc.

💻 Installation

Follow these steps to install and launch the interface:

1. Clone the project: git clone https://github.com/TheValll/Ollama-AI-Chat-Interface

2. Install dependencies: pip install -r requirements.txt

3. Run the Streamlit interface: ollama run [your_model]

   Note: You need to run your ollama before the streamlit interface

4. Run the Streamlit interface: streamlit run./ui.py --theme.base="light" --server.port=11434 --server.address=0.0.0.0 --server.headless=true

   Note: If the Ollama server URL is changed in the config.json file, the Streamlit port must match.

⚙️ Configuration

You can change the config.json file : json { "user_name": "user", "assistant_name": "Llama", "language": "en", "url": "http://localhost:11434/api/chat", "model": "llama3:8b" }

• user_name: User's name.
• assistant_name: Assistant's name.
• language: Interface language (default is "en", fr and en is available).
• url: URL of the Ollama server (default is http://localhost:11434/api/chat).
• model: Model used on Ollama (default is llama3:8b).

🚀 Using the Interface

The interface will be accessible via the IPv4 of the host computer + the default port 11434, from any device connected to the same network. Example : 192.168.1.2:11434

📂 Managing Categories

The user can manage multiple conversations called categories:

• Default Category: A default category is provided with a Full-Stack Software Developer prompt and cannot be deleted.
• Add a Category: Add a new category with a required name and prompt.
• Clear a Category: Clear the chat history of a category without deleting the initial prompt.
• Delete a Category: Delete a user-created category.

📝 History

All chat histories are stored in a history.json file. This file keeps a record of all conversations in each category.

❓ Why JSON?

The JSON format is chosen for history because it is easy to transport and interchangeable with other JSON files. It allows for simple and efficient management of conversation data.

🌐 Obtaining IPv4

To access the interface from other devices on the same network, you will need to know the IPv4 of the host computer. Here's how to obtain your IPv4:

On Windows:

Open Command Prompt (cmd). Type the following command: ipconfig

Look for the IPv4 address in the results under the network adapter in use (usually something like 192.168.x.x).

On macOS/Linux:

Open a terminal. Type the following command: ifconfig

Look for the IPv4 address in the results under the network interface in use (usually something like 192.168.x.x).

Use this IPv4 address followed by port 11434 to access the interface from other devices connected to the same network.

💡 Suggestions for Improvement

Feel free to propose ideas or improvements for this project. Your feedback is valuable for enhancing the user experience and features of the interface.

👨‍💻 Author

This project was developed by Valentin Massonniere.