ollama-gpt

ollama-gpt is a graphical interface tool designed to simplify the management and evolution of your ollama projects. The project is developed using GPT-o3, which drives continuous code optimization, enhanced documentation, and smarter testing processes. This innovative approach aims to transform the tool into a self-improving system that adapts to evolving development needs and industry trends.

Features

• One-Click Compilation: Compile the source code effortlessly using the integrated build system.
• Execution: Run the compiled executable with a single click for quick testing.
• Server Mode: Start the server (ollama serve) for deployment and debugging.
• Client Chat Interaction: Launch the chat interaction mode (ollama chat) using the default 1.5b model.
• Model Management: List and select supported models via an interactive interface.
• Interactive Command Input: Send real-time commands to interact with the running model.
• GPT-o3 Driven Development: Leverage GPT-o3 to continuously analyze and optimize the codebase, documentation, and testing procedures.

Prerequisites

• Essential Tools: Ensure you have the required tools installed for building and running the project (e.g., a compatible build system and command-line environment).
• ollama Source Code: The project directory should include the ollama folder and its related files.
• Optional Dependencies: While not strictly required, additional libraries such as PyQt5 can enhance the graphical interface experience.

Installation and Usage

1. Clone the Repository

git clone https://github.com/MarsDoge/ollama-gpt.git cd ollama-gpt

2. Install Optional Dependencies

For an enhanced GUI experience, install PyQt5 (optional): pip install PyQt5

3. Project Structure

The repository contains the following files and folders:

• llama_manager.py: The main program file providing the GUI and core functionalities.
• run.sh: A startup script that checks for the required environment and launches the main program.
• ollama/: A folder containing the necessary ollama source files and resources.

4. Launch the Application

Run the startup script:

• ./run.sh

5. User Guide

• Select Source Path: Click the "Browse" button to select the directory containing the ollama source code.
• Compile the Source: Click the "One-Click Compilation" button to compile the source code using the integrated build system. The "Run" button will become active upon successful compilation.
• Run the Program: Click the "Run" button to start the application.
• Server/Client Modes: Use the "Start Server" and "Start Client Chat Interaction" buttons to launch the respective modes.
• Model Management: Click the "List Supported Models" button to display available models. Select a model from the dropdown and click "Run Selected Model" to start an interactive session. Use the command input box to send real-time commands.

GPT-o3 Driven Development