AI A/B Testing: Which Prompt Performs Better?

AI A/B Testing: Which Prompt Performs Better?

💡 What You Will Learn

AI A/B Testing: Which Prompt Performs Better?

import random

prompts = {
    "A": "",
    "B": "3"
}

answers = []
for i in range(100):
    version = random.choice(["A", "B"])
    answer = call_llm(prompts[version], user_question)
    answers.append({"version": version, "answer": answer})

# 

def evaluate_answer(question, answer):
    prompt = f"""1-10
- 
- Complete
- 

{question}
{answer}
"""
    return int(call_llm(prompt))