holarissun/Prompt-OIRL
code for paper Query-Dependent Prompt Evaluation and Optimization with Offline Inverse Reinforcement Learning
This project helps AI developers and researchers improve how Large Language Models (LLMs) perform on specific tasks, especially for complex reasoning like arithmetic. It takes existing demonstration data of how different prompts perform with LLMs and learns an offline reward model to evaluate new prompts without needing to query the LLM directly. The output is an optimized prompt that is tailored to specific user queries, leading to better and more consistent LLM responses.
No commits in the last 6 months.
Use this if you are a machine learning engineer or researcher looking to efficiently optimize prompts for Large Language Models to get better, query-specific results without high computational costs.
Not ideal if you are an end-user simply looking for a better way to phrase your queries to an LLM without delving into model training or prompt engineering methodologies.
Stars
43
Forks
6
Language
Python
License
MIT
Category
Last pushed
Mar 20, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/prompt-engineering/holarissun/Prompt-OIRL"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
linshenkx/prompt-optimizer
一款提示词优化器，助力于编写高质量的提示词
Undertone0809/promptulate
🚀Lightweight Large language model automation and Autonomous Language Agents development...
CTLab-ITMO/CoolPrompt
Automatic Prompt Optimization Framework
microsoft/sammo
A library for prompt engineering and optimization (SAMMO = Structure-aware Multi-Objective...
Eladlev/AutoPrompt
A framework for prompt tuning using Intent-based Prompt Calibration