trustyai-explainability / lm-eval-tasks

Repo for lm-eval tasks built by the TrustyAI team

TrustyAI LM-Eval Tasks

This repo contains tasks for EleutherAI's LM Evaluation Harness built by the TrustyAI team. These tasks may be PoCs that are not yet ready to contribute upstream, or may simply be too TrustyAI-specific to warrant contribution to the main lm-eval repo.

Installation

Copy the tasks directory from this repo into the lm_eval/tasks directory, then build the Python module.