github/ColossalAI
1
0
Fork 0
mirror of https://github.com/hpcaitech/ColossalAI.git synced 2025-09-05 19:13:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

Revert "upgrade reward math verification"

Browse Source 
This reverts commit a6085ff676.
This commit is contained in:
YeAnbang
2025-05-07 10:50:41 +08:00
parent 2cd70cbcb1
commit d34115c5c4
1 changed files with 6 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
applications/ColossalChat/coati/distributed/reward/reward_fn.py
View File

@@ -1,5 +1,4 @@
import torch import torch
from math_verify import parse, verify
from .reward_utils import extract_boxed_solution, extract_solution, validate_response_structure from .reward_utils import extract_boxed_solution, extract_solution, validate_response_structure
@@ -36,7 +35,11 @@ def math_reward_fn(input_ids, gt_answer, response_idx, **kwargs):
format_acc += 1 format_acc += 1
# Check answer accuracy, answer is considered correct if the answer is correct and the format is valid # Check answer accuracy, answer is considered correct if the answer is correct and the format is valid
if format_valid and final_answer is not None and verify(parse(gt_answer.strip()), parse(final_answer.strip())): if (
format_valid
and final_answer is not None
and gt_answer.strip().replace(" ", "").lower() == final_answer.strip().replace(" ", "").lower()
):
ans_acc += 1 ans_acc += 1
reward += acc_score reward += acc_score
@@ -88,7 +91,7 @@ def boxed_math_reward_fn(input_ids, gt_answer, response_idx, **kwargs):
reward += format_score reward += format_score
# Check answer accuracy, answer is considered correct if the answer is correct and the format is valid # Check answer accuracy, answer is considered correct if the answer is correct and the format is valid
if format_valid and final_answer is not None and verify(parse(gt_answer.strip()), parse(final_answer.strip())): if format_valid and final_answer is not None and gt_answer.strip().lower() == final_answer.strip().lower():
ans_acc += 1 ans_acc += 1
reward += acc_score reward += acc_score