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 7475a20700
commit ce0a6c4878
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
104 changed files with 549 additions and 146 deletions

View file

@ -24,6 +24,9 @@ Return True if the following two strings are isomorphic, or False otherwise:
"""
DATASET_NAME = "isomorphic_strings"
@dataclass
class IsomorphicStringsConfig:
"""Configuration for Isomorphic Strings dataset generation"""
@ -107,6 +110,8 @@ class IsomorphicStringsDataset(ProceduralDataset):
"question": QUESTION_TEMPLATE.format(s=s, t=t),
"answer": str(answer),
"metadata": {
"source_dataset": DATASET_NAME,
"source_index": idx,
"words": [s, t],
"solution": answer,
"solvable": solvable,
@ -134,4 +139,4 @@ class IsomorphicStringsCurriculum(BaseCurriculum):
)
register_dataset("isomorphic_strings", IsomorphicStringsDataset, IsomorphicStringsConfig, IsomorphicStringsCurriculum)
register_dataset(DATASET_NAME, IsomorphicStringsDataset, IsomorphicStringsConfig, IsomorphicStringsCurriculum)