Generative AI Coding Benchmark

Generated by ChatGPT for CS 8395

Overview

This repository contains a set of 100 unique coding problems designed to assess various aspects of Generative AI development. These problems focus on areas like text and image generation, sequence-to-sequence models, transfer learning, evaluation metrics, and many more.

What's New and Novel?

Traditional coding assessments often focus on algorithmic skills, data structures, or web development tasks. This coding benchmark is unique in its focus on Generative AI, a rapidly evolving field with wide-ranging applications. Each problem is specifically designed to assess a particular aspect of Generative AI, such as building generative models, text generation, image synthesis, and even tackling ethical considerations in AI.

What Does It Assess?

The benchmark is designed to assess the following:

• Understanding and implementation of generative models
• Ability to generate text and images using various techniques
• Skills in sequence-to-sequence models and their applications
• Understanding and implementation of transfer learning
• Ability to design custom evaluation metrics
• Understanding of attention mechanisms and their debugging
• Skills in data preparation and preprocessing
• Understanding of optimization techniques specific to generative models

How to Run the Benchmark

Prerequisites

• Python 3.x
• Necessary Python packages (e.g., TensorFlow, PyTorch, NumPy)

Steps

1. Clone the Repository: Clone this repository to your local machine.

   git clone https://github.com/your_username/generative_ai_coding_benchmark.git

2. Navigate to the Directory: Go to the directory containing the problems.

   cd generative_ai_coding_benchmark

3. Read the Problems: Open the generative_ai_unique_descriptive_coding_problems.md file to read through the problems.

4. Implement Solutions: For each problem, there is a descriptive function header to guide your implementation. Replace the parameters in the function header with the actual parameters required for each problem.

5. Run Tests (Optional): If there are any tests provided or any grading criteria, run them to assess your solution.

6. Submit: Once you've implemented the solutions, you can submit them according to the guidelines of your assessment.

Contributing

Feel free to contribute to this benchmark by adding more problems, tests, or grading criteria. Pull requests are welcome.

License

This project is licensed under the MIT License.

Acknowledgements

This benchmark was generated using OpenAI's GPT-4 and is designed to serve as a comprehensive assessment tool for Generative AI skills.