open-thought/reasoning-gym
1
0
Fork 0
mirror of https://github.com/open-thought/reasoning-gym.git synced 2026-04-29 17:35:16 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

updated spell back

Browse source
This commit is contained in:
joesharratt1229 2025-03-26 19:53:27 +00:00
parent 16919223be
commit f50c7221ac
1 changed files with 13 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

17
reasoning_gym/algorithmic/spell_backward.py
View file

@ -68,12 +68,21 @@ class SpellBackwardDataset(ProceduralDataset):
expected_answer = entry["answer"]
if isinstance(answer, str):
try:
if expected_answer.lower() == answer.lower():
expected_answer = expected_answer.lower()
answer = answer.lower()
if expected_answer == answer:
reward = 1.0
elif sorted(expected_answer.lower()) == sorted(answer.lower()):
reward = 0.2
else:
reward = 0.05
answer_len = len(answer)
for i in range(len(expected_answer)):
if (i < len(expected_answer) and i < len(answer)) and expected_answer[i] == answer[i]:
if expected_answer[i] == answer[i]:
reward += 1 / answer_len
else:
continue
else:
break
except:
reward = 0.0
return reward