Minor question template & score_answer improvements (#261)

* math prompt improvements
* ignore brackets in complex_arithmetic results
* improve additional instruction in prompt of polynomial_equations
* more strict tests for score_answer in polynomial_equations
* simplify special reward handling
* fix test_intermediate_integration
* fix sokoban dataset
* add common dataset score_answer consistency test
This commit is contained in:
Andreas Köpf 2025-03-04 21:55:09 +01:00 committed by GitHub
parent 061282e373
commit 5d7fbac0ad
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
106 changed files with 403 additions and 507 deletions

View file

@ -160,17 +160,15 @@ class BitwiseArithmeticDataset(ProceduralDataset):
Returns:
float: 1.0 if the user's answer is correct; otherwise, 0.01 unless no answer is provided, in which case 0.
"""
if answer is None:
return 0.0
if isinstance(answer, str):
try:
solved = verify_solution(entry["metadata"]["problem"], answer)
if solved:
return 1.0
except Exception:
pass
try:
solved = verify_solution(entry["metadata"]["problem"], answer)
if solved:
return 1.0
except Exception:
return 0.01
return 0.01
return 0.0
# Register the dataset with the factory.