Datasets:
Tasks:
Text Generation
Formats:
parquet
Sub-tasks:
language-modeling
Languages:
English
Size:
< 1K
License:
Update dataset (part 3)
Browse files
harbor/sources/hooks-v4/tasks.py
CHANGED
|
@@ -146,6 +146,7 @@ async def _judge_diff(name: str, weight: float, diff: str) -> SubScore:
|
|
| 146 |
answer=diff[:100_000],
|
| 147 |
criteria=list(JUDGE["criteria"]),
|
| 148 |
question=JUDGE["question"],
|
|
|
|
| 149 |
)
|
| 150 |
except Exception as exc: # noqa: BLE001 - judge failure must not kill grading
|
| 151 |
return SubScore(
|
|
|
|
| 146 |
answer=diff[:100_000],
|
| 147 |
criteria=list(JUDGE["criteria"]),
|
| 148 |
question=JUDGE["question"],
|
| 149 |
+
model=os.environ.get("JUDGE_MODEL", "claude-sonnet-5-5"),
|
| 150 |
)
|
| 151 |
except Exception as exc: # noqa: BLE001 - judge failure must not kill grading
|
| 152 |
return SubScore(
|
harbor/sources/sequelize-v27/tasks.py
CHANGED
|
@@ -134,6 +134,7 @@ async def _judge_diff(name: str, weight: float, diff: str) -> SubScore:
|
|
| 134 |
answer=diff[:100_000],
|
| 135 |
criteria=list(JUDGE["criteria"]),
|
| 136 |
question=JUDGE["question"],
|
|
|
|
| 137 |
)
|
| 138 |
except Exception as exc: # noqa: BLE001 - judge failure must not kill grading
|
| 139 |
return SubScore(
|
|
|
|
| 134 |
answer=diff[:100_000],
|
| 135 |
criteria=list(JUDGE["criteria"]),
|
| 136 |
question=JUDGE["question"],
|
| 137 |
+
model=os.environ.get("JUDGE_MODEL", "claude-sonnet-5-5"),
|
| 138 |
)
|
| 139 |
except Exception as exc: # noqa: BLE001 - judge failure must not kill grading
|
| 140 |
return SubScore(
|