anessbelbati/jev-rerank-bench on GitHub
- Stars
- 7
- Forks
- 0
- Language
- Python
- License
- MIT
- Last push
- 23 Sep 2026
Read from GitHub on . Counts change daily.
How it uses Jev
# JEV_URL / JEV_MODEL let the same request builders run against a local server that speaks the same protocol
# (used for the open-weight Open-Jev adapters); JEV_GPU_RATE (USD per hour) then bills GPU seconds instead of tokens.
URL = os.environ.get("JEV_URL", "https://api.typesafe.ai/v1/systemone")
MODEL = os.environ.get("JEV_MODEL", "jev-latest")
RUBRIC = ["The passage is off-topic for the query.",
"The passage is on a related topic but does not supply what the query asks for.",
"The passage partly supplies the information needed to answer or verify the query.",
"The passage fully supplies the information needed to answer or verify the query."]
def _headers() -> dict:
key = os.environ.get("JEV_API_KEY", "") if os.environ.get("JEV_URL") else env("JEV_API_KEY")
return {"Authorization": f"Bearer {key}", "Content-Type": "application/json", "Accept": "application/json"}PrimitivesNoul
Use casesSemantic reranking
