Datasets:
Register krama_full in bundle manifest
Browse files- dataset_manifest.json +51 -0
dataset_manifest.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"schema_version": "evo_eval.dataset_bundle.v1",
|
| 3 |
+
"benchmarks": {
|
| 4 |
+
"bird": {
|
| 5 |
+
"data_dir": "bird",
|
| 6 |
+
"task_count": 1534,
|
| 7 |
+
"prediction_protocol": "code:prediction",
|
| 8 |
+
"task_id_protocol": "one_based_1_to_1534",
|
| 9 |
+
"dataset_schema": "evo_eval.dataset.v1"
|
| 10 |
+
},
|
| 11 |
+
"dacomp": {
|
| 12 |
+
"data_dir": "dacomp",
|
| 13 |
+
"task_count": 100,
|
| 14 |
+
"prediction_protocol": "document:prediction",
|
| 15 |
+
"task_id_protocol": "one_based_1_to_100",
|
| 16 |
+
"language": "zh",
|
| 17 |
+
"paired_benchmark": "dacomp-en",
|
| 18 |
+
"reference_protocol": "dacomp_document_judge.v1",
|
| 19 |
+
"dataset_schema": "evo_eval.dataset.v1"
|
| 20 |
+
},
|
| 21 |
+
"dacomp-en": {
|
| 22 |
+
"data_dir": "dacomp-en",
|
| 23 |
+
"task_count": 100,
|
| 24 |
+
"prediction_protocol": "document:prediction",
|
| 25 |
+
"task_id_protocol": "one_based_1_to_100",
|
| 26 |
+
"language": "en",
|
| 27 |
+
"paired_benchmark": "dacomp",
|
| 28 |
+
"reference_protocol": "dacomp_document_judge.v1",
|
| 29 |
+
"dataset_schema": "evo_eval.dataset.v1"
|
| 30 |
+
},
|
| 31 |
+
"sql_feature": {
|
| 32 |
+
"data_dir": "sql_feature",
|
| 33 |
+
"task_count": 145,
|
| 34 |
+
"prediction_protocol": "code:prediction",
|
| 35 |
+
"task_id_protocol": "one_based_1_to_145",
|
| 36 |
+
"language": "zh",
|
| 37 |
+
"reference_protocol": "sql_feature_judge.v1",
|
| 38 |
+
"dataset_schema": "evo_eval.dataset.v1"
|
| 39 |
+
},
|
| 40 |
+
"krama_full": {
|
| 41 |
+
"data_dir": "krama_full",
|
| 42 |
+
"task_count": 104,
|
| 43 |
+
"prediction_protocol": "document:prediction",
|
| 44 |
+
"task_id_protocol": "one_based_1_to_104",
|
| 45 |
+
"subtask_count": 631,
|
| 46 |
+
"source_benchmark": "KramaBench",
|
| 47 |
+
"reference_protocol": "krama_metric.v1",
|
| 48 |
+
"dataset_schema": "evo_eval.dataset.v1"
|
| 49 |
+
}
|
| 50 |
+
}
|
| 51 |
+
}
|