evals: eval crate overhaul, simplification and performance improvements

This commit is contained in:
Per Stark
2026-06-17 19:23:11 +02:00
parent adc04d8c6d
commit fb51a8b55f
53 changed files with 2852 additions and 1831 deletions
+2 -1
View File
@@ -1,4 +1,4 @@
default_dataset: squad-v2
default_dataset: beir
datasets:
- id: squad-v2
label: "SQuAD v2.0"
@@ -45,6 +45,7 @@ datasets:
description: "Balanced slice across FEVER, FiQA, HotpotQA, NFCorpus, Quora, TREC-COVID, SciFact, NQ-BEIR"
limit: 600
corpus_limit: 6000
negative_multiplier: 9.0
seed: 0x5eed2025
- id: fever
label: "FEVER (BEIR)"