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

@ -17,6 +17,9 @@ Reverse this list of words: {words}
"""
DATASET_NAME = "word_sequence_reversal"
@dataclass
class WordSequenceReversalConfig:
"""Configuration for word sequence reversal task generation"""
@ -63,6 +66,8 @@ class WordSequenceReversalDataset(ProceduralDataset):
"question": f"{QUESTION_TEMPLATE.format(words=words_str)}",
"answer": answer,
"metadata": {
"source_dataset": DATASET_NAME,
"source_index": idx,
"num_words": num_words,
"words": words,
"difficulty": {
@ -89,6 +94,4 @@ class WordSequenceReversalCurriculum(BaseCurriculum):
)
register_dataset(
"word_sequence_reversal", WordSequenceReversalDataset, WordSequenceReversalConfig, WordSequenceReversalCurriculum
)
register_dataset(DATASET_NAME, WordSequenceReversalDataset, WordSequenceReversalConfig, WordSequenceReversalCurriculum)