open-thought/reasoning-gym
1
0
Fork 0
mirror of https://github.com/open-thought/reasoning-gym.git synced 2026-04-19 12:58:07 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

adapted score answer

Browse source
This commit is contained in:
joesharratt1229 2025-02-08 11:37:25 +00:00
parent ab33c2bcbb
commit a1494c4e5b
1 changed files with 9 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

13
reasoning_gym/arc/rearc.py
View file

@ -3,7 +3,13 @@ from random import Random
from typing import Any, Callable, Dict, Optional
from ..factory import ProceduralDataset, register_dataset
from .rearc_board_format import BoardFormattingOptions, default_board_format_opts, format_board, format_board_pair
from .rearc_board_format import (
BoardFormattingOptions,
default_board_format_opts,
format_board,
format_board_pair,
parse_board,
)
from .rearc_utils import generators, verifiers
from .rearc_utils.dsl import *
from .rearc_utils.utils import *
@ -166,9 +172,8 @@ class ReArcDataset(ProceduralDataset):
reward = 0.0
if answer is not None:
try:
task_id = metadata["task_id"]
verifier = self._verifiers[task_id]
if verifier(metadata["input"]) == metadata["output"]:
formatted_answer = parse_board(answer, self.board_format_opts)
if formatted_answer == metadata["output"]:
reward = 1.0
else:
reward = 0.05