README.md

PromptGenCLI

This repository demonstrates how to interact with OpenAI's API to generate text-based responses using customizable inference parameters. The program allows developers to send a prompt to a selected model, returning a detailed response based on specific input parameters like temperature and top-p, which influence the response variability and output length.

Features

• Model Customization: Choose from different OpenAI models to generate text-based completions.
• Inference Parameters: Control the variability of the response using temperature and top-p, which adjust how creative or focused the model should be.
• Token Length: Supports up to 2000 tokens per response, controlling the length and complexity of generated text.
• Command-Line Interface: Accepts input model and prompt through command-line arguments for flexible usage.
• Stop Sequences: Optional inference parameters (like Stop Sequences) can be used to fine-tune when and how the response should end, preventing unwanted or repetitive outputs.

Inference Parameters Overview

Inference parameters customize how the foundation model behaves. These include:

• Temperature: Controls the randomness of the model’s output. A higher temperature (e.g., 1.0) makes the output more diverse, while a lower value (e.g., 0) makes it more deterministic.

• Top-p: Adjusts how much of the output is focused on the highest probability tokens. Lower top-p values restrict the output to more likely tokens.

• Max Tokens: Determines the maximum length of the output, where tokens can be either a word or part of a word.

• Stop Sequences: Help in controlling the end of the response, stopping the output when a certain phrase or token sequence is encountered.

Usage

Prerequisites

1. Clone the repository:

   git clone https://github.com/yourusername/PromptGenCLI.git

2. Install dependencies:

   pip install openai python-dotenv

3. Create a .env file and add your OpenAI API key:

   OPENAI_API_KEY=your-api-key-here

Running the Program

The program accepts two arguments:

1. The OpenAI model (e.g., gpt-3.5-turbo)
2. The prompt for the model to process.

To run the program:

python params.py gpt-3.5-turbo "What is the meaning of life?"

Example Output

The meaning of life can vary greatly depending on personal beliefs, culture, and experiences...

Customization

Modify the temperature, top_p, or max_tokens to change the behavior and length of responses:

• temperature: Set between 0 and 1 for more varied or focused responses.
• top_p: Set between 0 and 1 to adjust the probability distribution of token selection.

License

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

---