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

@ -49,6 +49,8 @@ Tip: Here is a reference code snippet for this question. You can refer to this c
{3}
"""
DATASET_NAME = "codeio"
@dataclass
class CodeIOConfig:
@ -117,7 +119,12 @@ class CodeIODataset(ProceduralDataset):
return {
"question": question,
"answer": solution,
"metadata": {"input_data": input_data, "output_data": output_data},
"metadata": {
"source_dataset": DATASET_NAME,
"source_index": idx,
"input_data": input_data,
"output_data": output_data,
},
}
def _json_to_tree(self, data, label="root"):
@ -231,4 +238,4 @@ class CodeIODataset(ProceduralDataset):
# Register the dataset
register_dataset("codeio", CodeIODataset, CodeIOConfig)
register_dataset(DATASET_NAME, CodeIODataset, CodeIOConfig)