Ollama Prompt Tool

This repository provides a simple command-line interface (CLI) tool for interacting with language models using the ollama Python library. Users can input prompts, receive model-generated responses, and maintain a history of interactions for context in subsequent queries.

Features

• User-friendly CLI interface
• Continuous interaction with a language model
• Context preservation for better response relevance
• Handles errors gracefully

---

Prerequisites

1. Install Python

Ensure you have Python 3.8 or higher installed. You can download it from the official Python website.

2. Install Required Libraries

Use pip to install the required dependencies:

pip install ollama

3. Compatible Models

This tool works with models available through ollama, such as vicuna:7b. Ensure you have the desired model downloaded and ready for use.

---

Installation

1. Clone this repository:

git clone https://github.com/yourusername/ollama-prompt-tool.git
cd ollama-prompt-tool

2. Install dependencies:

pip install -r requirements.txt

---

Usage

Run the script from your terminal:

python main.py

Interacting with the Model

1. Enter a prompt when prompted.
2. Press Enter to submit the prompt.
3. The model's response will be displayed.
4. To exit, press Enter without typing a prompt.

---

Code Explanation

import ollama

# Initialize a list to store responses
response_history = []

while True:
    # Get user input for the prompt (display in green)
    prompt = input("\033[92m\nEnter your prompt (or press Enter to exit): \033[0m")
    if prompt == '':
        break
    
    # Combine previous responses as context
    context = "\n".join(response_history)
    full_prompt = context + "\n" + prompt if context else prompt
    
    try:
        # Generate a response using the model
        response = ollama.generate(model='vicuna:7b', prompt=full_prompt)
        
        # Store the response in the history
        response_history.append(response.response)
        
        # Print the latest response in cyan
        print("\033[96m\n" + response.response + "\033[0m")
    except Exception as e:
        # Print the error in red
        print("\033[91m\nAn error occurred:\033[0m", "\033[91m" + str(e) + "\033[0m")

• response_history: Stores previous responses to maintain context.
• ollama.generate: Generates a response using the specified model.
• Error Handling: Captures and displays errors in red text.

---

Contributing

Contributions are welcome! Feel free to submit issues or pull requests to improve the tool.

---

License

This project is licensed under the MIT License. See the LICENSE file for details.