GitLab

# Conflicts:
#	lm_eval/api/task.py

[Refactor] float16 MPS works in torch nightly

Merge branch 'big-refactor' of https://github.com/EleutherAI/lm-evaluation-harness into better-benchmark

[Refactor] Fix MGSM

[Refactor] Additions to docs

Add pyproject.toml

[Refactor] Add top-level import

[Refactor] mgsm

Update README.md

Crossed WMT off.

[Refactor] consolidated tasks tests