run · run_ady8ctyw9gfcde2v · filed 2026-07-11 06:12:15
Share

BM25 only (property text)

Grader output

free-form JSON · this task's own
0.560grader.score28 of 50 cases passed
score
0.560
total cases
50
passed cases
28
skipped cases
0
latency total
23.00 s

Fields here are the task author's — trap neither interprets nor validates them.

Run frame

written by trap · always present
cases
50
duration total
23.00s
cost
engine
bm25 → qwen-plus

environment · self-reported, not validated

strategy
BM25 single-path retrieval over property text (Elasticsearch) + qwen-plus
answer_model
qwen-plus
retrieval_config
bm25
per-case results · all 50, in case_id order
discuss this task →