← All repos

ab_challenge_eval

Evaluator for the A::B Prompting Challenge

GitHub
Browse cluster: Cluster 89 →
1commits
1contributors
1languages

Tech stack & purpose

ab_challenge_eval is an evaluator tool for the A::B Prompting Challenge that tests and compares system prompts. The project is built with Node.js and JavaScript, requiring users to configure API keys for Anthropic or OpenAI and provide their own system prompt and model selection. To run the evaluator, users install dependencies via npm and execute the main script with their participant name.

Languages

JavaScript
100.0%

Contributors