Files

Dominik Kundel 243a1b0276 Initial commit

Co-authored-by: Zhuohan Li <zhuohan@openai.com>
Co-authored-by: Maratyszcza <marat@openai.com>
Co-authored-by: Volodymyr Kyrylov <vol@wilab.org.ua>

2025-08-05 08:19:49 -07:00

__init__.py

Initial commit

2025-08-05 08:19:49 -07:00

__main__.py

Initial commit

2025-08-05 08:19:49 -07:00

abcd_grader.py

Initial commit

2025-08-05 08:19:49 -07:00

aime_eval.py

Initial commit

2025-08-05 08:19:49 -07:00

chat_completion_sampler.py

Initial commit

2025-08-05 08:19:49 -07:00

gpqa_eval.py

Initial commit

2025-08-05 08:19:49 -07:00

healthbench_eval.py

Initial commit

2025-08-05 08:19:49 -07:00

README.md

Initial commit

2025-08-05 08:19:49 -07:00

report.py

Initial commit

2025-08-05 08:19:49 -07:00

responses_sampler.py

Initial commit

2025-08-05 08:19:49 -07:00

types.py

Initial commit

2025-08-05 08:19:49 -07:00

README.md

`gpt_oss.evals`

This module is a reincarnation of simple-evals adapted for gpt-oss. It lets you run GPQA and HealthBench against a runtime that supports Responses API on localhost:8080/v1.