fix(envs): Add source dataset and index to metadata (#388)

* add source dataset and index to metadata

* fix typo

* fix coach class and its test
This commit is contained in:
Zafir Stojanovski 2025-03-20 12:12:14 +01:00 committed by GitHub
parent c6d01541aa
commit 4c47527130
104 changed files with 549 additions and 146 deletions

View file

@ -5,6 +5,8 @@ from typing import Any, Optional
from ..coaching import BaseCurriculum, ScalarAttributeDefinition
from ..factory import ProceduralDataset, register_dataset
DATASET_NAME = "bitwise_arithmetic"
@dataclass
class BitwiseArithmeticConfig:
@ -155,7 +157,12 @@ class BitwiseArithmeticDataset(ProceduralDataset):
return {
"question": problem_str,
"answer": answer,
"metadata": {"problem": problem, "difficulty": {"difficulty": self.config.difficulty}},
"metadata": {
"source_dataset": DATASET_NAME,
"source_index": idx,
"problem": problem,
"difficulty": {"difficulty": self.config.difficulty},
},
}
def score_answer(self, answer: Optional[str], entry: dict[str, Any]) -> float:
@ -193,4 +200,4 @@ class BitwiseArithmeticCurriculum(BaseCurriculum):
# Register the dataset with the factory.
register_dataset("bitwise_arithmetic", BitwiseArithmeticDataset, BitwiseArithmeticConfig, BitwiseArithmeticCurriculum)
register_dataset(DATASET_NAME, BitwiseArithmeticDataset, BitwiseArithmeticConfig, BitwiseArithmeticCurriculum)