open-thought/reasoning-gym
1
0
Fork 0
mirror of https://github.com/open-thought/reasoning-gym.git synced 2026-04-30 17:40:45 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

added preappend token

Browse source
This commit is contained in:
joesharratt1229 2025-04-01 16:28:04 +00:00
parent 4b9c155cef
commit 8d0e7db204
2 changed files with 6 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

4
training/rewards/reward.py
View file

@ -60,6 +60,10 @@ def cosine_scaled_reward(solution_str, scaling_factor, **kwargs):
@reward_registry.register("format")
def compute_format_reward(solution_str: str, scaling_factor: float = 0.2, **kwargs) -> float:
"""Reward use of exactly one correctly structured <think> and <answer> block."""
preappend_thinking_token = kwargs.get("preappend_thinking_token", False)
if preappend_thinking_token:
solution_str = "<think>" + solution_str
pattern = r"\s*<think>.*?</think>\s*<answer>.*?</answer>"
if not re.match(pattern, solution_str, re.DOTALL):
return 0.0