promptfletcher 0.1.0

A Python library for auto-prompt engineering and optimization for LLMs.

PromptFletcher 🚀

A Python library for auto-prompt engineering and optimization for LLMs.

---

What is PromptFletcher?

PromptFletcher is a lightweight and powerful Python library designed for:
Refining & optimizing prompts for LLMs
Context-aware prompt tuning for better responses
Heuristic-based evaluation to rank prompts
Integrating with open-source LLMs (GPT-Neo, GPT-J, BLOOM)

---

Installation

From PyPI (Coming Soon)

pip install promptfletcher

From GitHub

pip install git+https://github.com/Vikhram-S/PromptFletcher.git

---

Quick Start

Import & Initialize

from promptfletcher import AutoPromptEngineer

engineer = AutoPromptEngineer()

Define Context & Prompt

context = "We are exploring ways to enhance prompt engineering for LLMs."
initial_prompt = "How can I improve my AI-generated responses?"

Optimize the Prompt

refined_prompt = engineer.refine_prompt(initial_prompt, context)
print("Refined Prompt:", refined_prompt)

---

Features

Automated Prompt Refinement – Uses NLP techniques to improve prompt clarity.
LLM Response Evaluation – Integrates with open-source models like GPT-Neo & BLOOM.
Contextual Understanding – Ensures prompts align with relevant topics.
Lightweight & Fast – Minimal dependencies, designed for efficiency.

---

API Reference

AutoPromptEngineer Class

refine_prompt(prompt: str, context: str, iterations: int = 3) -> str

Refines a given prompt based on context and heuristic scoring.

engineer.refine_prompt("How do I make my AI-generated text more accurate?", "LLM optimization")

evaluate_prompt(prompt: str, context: str) -> float

Assigns a heuristic score to a prompt based on clarity and relevance.

score = engineer.evaluate_prompt("Tell me about AI safety?", "Machine Learning Ethics")
print("Prompt Score:", score)

---

Dependencies

• transformers>=4.30.0
• torch>=2.0.0
• numpy>=1.21.0
• regex>=2023.3.23

Install dependencies manually:

pip install -r requirements.txt

---

License

PromptFletcher is licensed under the MIT License – free to use, modify, and distribute.

---

Contributing

We welcome contributions!

1. Fork the repository
2. Create a feature branch (git checkout -b feature-new)
3. Commit changes & push (git push origin feature-new)
4. Open a Pull Request

---

Contact & Support

• GitHub Issues: Report Bugs
• Email: [e-mail]

If you find this useful, give us a star on GitHub!

---