Watch
1
0
Fork
You've already forked Seyyed_arc
0
forked from hesabix/arc
Seyyed_arc/docs/ai-agent/evals/tool-hybrid/latest.json
2026-08-21 00:15:57 +00:00

95 lines
No EOL
2.3 KiB
JSON

{
"hybrid": [
{
"queries": 443,
"recall": 0.9458,
"mrr": 0.7845,
"ndcg": 0.8023,
"precision": 0.1808,
"avg_k": 10.0,
"avg_initial_schema_tokens": 2754,
"avg_schema_tokens_per_request": 14462,
"k": 10,
"weights": "lexical_dominant",
"delete_family_false_positives": 1,
"delete_family_queries": 20
},
{
"queries": 443,
"recall": 0.9639,
"mrr": 0.7858,
"ndcg": 0.8091,
"precision": 0.1251,
"avg_k": 15.0,
"avg_initial_schema_tokens": 4025,
"avg_schema_tokens_per_request": 20944,
"k": 15,
"weights": "lexical_dominant",
"delete_family_false_positives": 1,
"delete_family_queries": 20
},
{
"queries": 443,
"recall": 0.9458,
"mrr": 0.7847,
"ndcg": 0.7989,
"precision": 0.1779,
"avg_k": 10.0,
"avg_initial_schema_tokens": 2739,
"avg_schema_tokens_per_request": 14325,
"k": 10,
"weights": "balanced",
"delete_family_false_positives": 1,
"delete_family_queries": 20
},
{
"queries": 443,
"recall": 0.9616,
"mrr": 0.7857,
"ndcg": 0.8068,
"precision": 0.124,
"avg_k": 15.0,
"avg_initial_schema_tokens": 4013,
"avg_schema_tokens_per_request": 20872,
"k": 15,
"weights": "balanced",
"delete_family_false_positives": 1,
"delete_family_queries": 20
},
{
"queries": 443,
"recall": 0.9413,
"mrr": 0.7838,
"ndcg": 0.7956,
"precision": 0.1758,
"avg_k": 10.0,
"avg_initial_schema_tokens": 2722,
"avg_schema_tokens_per_request": 14221,
"k": 10,
"weights": "semantic_dominant",
"delete_family_false_positives": 1,
"delete_family_queries": 20
},
{
"queries": 443,
"recall": 0.9571,
"mrr": 0.7853,
"ndcg": 0.8043,
"precision": 0.1228,
"avg_k": 15.0,
"avg_initial_schema_tokens": 3956,
"avg_schema_tokens_per_request": 20556,
"k": 15,
"weights": "semantic_dominant",
"delete_family_false_positives": 1,
"delete_family_queries": 20
}
],
"best": {
"weights": "lexical_dominant",
"k": 15,
"recall": 0.9639,
"mrr": 0.7858,
"ndcg": 0.8091
}
}