CAEL-0.3 / evaluation /sql_parity.json
vynly's picture
Publish CAEL 0.3: frozen helper candidate, Wikipedia retrieval and no-image bundle
abc20fb verified
Raw History Blame Contribute Delete
2.38 kB
[
{
"task": "hellaswag",
"identity": "24",
"fresh_sql_prediction": 3,
"cached_prediction": 3,
"route": "language+evidence"
},
{
"task": "hellaswag",
"identity": "13364",
"fresh_sql_prediction": 2,
"cached_prediction": 2,
"route": "language"
},
{
"task": "hellaswag",
"identity": "50657",
"fresh_sql_prediction": 0,
"cached_prediction": 0,
"route": "language"
},
{
"task": "arc_easy",
"identity": "Mercury_417466",
"fresh_sql_prediction": 0,
"cached_prediction": 0,
"route": "language+evidence"
},
{
"task": "arc_easy",
"identity": "Mercury_7235603",
"fresh_sql_prediction": 2,
"cached_prediction": 2,
"route": "language+evidence"
},
{
"task": "arc_easy",
"identity": "VASoL_2009_5_34",
"fresh_sql_prediction": 0,
"cached_prediction": 0,
"route": "language+evidence"
},
{
"task": "arc_challenge",
"identity": "Mercury_7175875",
"fresh_sql_prediction": 1,
"cached_prediction": 1,
"route": "language+evidence"
},
{
"task": "arc_challenge",
"identity": "Mercury_SC_400839",
"fresh_sql_prediction": 3,
"cached_prediction": 3,
"route": "language+evidence"
},
{
"task": "arc_challenge",
"identity": "MEA_2013_8_15",
"fresh_sql_prediction": 2,
"cached_prediction": 2,
"route": "language+evidence"
},
{
"task": "piqa",
"identity": "0",
"fresh_sql_prediction": 1,
"cached_prediction": 1,
"route": "language"
},
{
"task": "piqa",
"identity": "919",
"fresh_sql_prediction": 1,
"cached_prediction": 1,
"route": "language+evidence"
},
{
"task": "piqa",
"identity": "1837",
"fresh_sql_prediction": 1,
"cached_prediction": 1,
"route": "language+evidence"
},
{
"task": "arithmark3",
"identity": "0",
"fresh_sql_prediction": 1,
"cached_prediction": 1,
"route": "exact-arithmetic"
},
{
"task": "arithmark3",
"identity": "500",
"fresh_sql_prediction": 2,
"cached_prediction": 2,
"route": "exact-arithmetic"
},
{
"task": "arithmark3",
"identity": "999",
"fresh_sql_prediction": 3,
"cached_prediction": 3,
"route": "exact-arithmetic"
}
]