[example] change dapo verify

examples/qwen3_14b_dapo17k_dapo.sh

@@ -10,7 +10,7 @@ python3 -m verl.trainer.main \
     config=examples/config.yaml \
     data.train_files=Saigyouji-Yuyuko1000/dapo17k@train \
     data.val_files=Saigyouji-Yuyuko1000/dapo17k@test \
-    data.format_prompt=./examples/format_prompt/math.jinja \
+    data.format_prompt=./examples/format_prompt/dapo.jinja \
     data.max_prompt_length=2048 \
     data.max_response_length=20480 \
     data.rollout_batch_size=512 \

examples/reward_function/dapo.py

@@ -12,14 +12,134 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+import re
 from typing import Any, Dict, List
 
-from mathruler.grader import extract_boxed_content, grade_answer
+
+# Constants for normalization
+SUBSTITUTIONS = [
+    ("an ", ""),
+    ("a ", ""),
+    (".$", "$"),
+    ("\\$", ""),
+    (r"\ ", ""),
+    (" ", ""),
+    ("mbox", "text"),
+    (",\\text{and}", ","),
+    ("\\text{and}", ","),
+    ("\\text{m}", "\\text{}"),
+]
+
+REMOVED_EXPRESSIONS = [
+    "square",
+    "ways",
+    "integers",
+    "dollars",
+    "mph",
+    "inches",
+    "hours",
+    "km",
+    "units",
+    "\\ldots",
+    "sue",
+    "points",
+    "feet",
+    "minutes",
+    "digits",
+    "cents",
+    "degrees",
+    "cm",
+    "gm",
+    "pounds",
+    "meters",
+    "meals",
+    "edges",
+    "students",
+    "childrentickets",
+    "multiples",
+    "\\text{s}",
+    "\\text{.}",
+    "\\text{\ns}",
+    "\\text{}^2",
+    "\\text{}^3",
+    "\\text{\n}",
+    "\\text{}",
+    r"\mathrm{th}",
+    r"^\circ",
+    r"^{\circ}",
+    r"\;",
+    r",\!",
+    "{,}",
+    '"',
+    "\\dots",
+]
+
+
+def normalize_final_answer(final_answer: str) -> str:
+    """Normalize a final answer to a quantitative reasoning question.
+
+    Args:
+        final_answer: The answer string to normalize
+
+    Returns:
+        Normalized answer string
+    """
+    final_answer = final_answer.split("=")[-1]
+
+    # Apply substitutions and removals
+    for before, after in SUBSTITUTIONS:
+        final_answer = final_answer.replace(before, after)
+    for expr in REMOVED_EXPRESSIONS:
+        final_answer = final_answer.replace(expr, "")
+
+    # Extract and normalize LaTeX math
+    final_answer = re.sub(r"(.*?)(\$)(.*?)(\$)(.*)", "$\\3$", final_answer)
+    final_answer = re.sub(r"(\\text\{)(.*?)(\})", "\\2", final_answer)
+    final_answer = re.sub(r"(\\textbf\{)(.*?)(\})", "\\2", final_answer)
+    final_answer = re.sub(r"(\\overline\{)(.*?)(\})", "\\2", final_answer)
+    final_answer = re.sub(r"(\\boxed\{)(.*)(\})", "\\2", final_answer)
+
+    # Normalize shorthand TeX:
+    #  \fracab -> \frac{a}{b}
+    #  \frac{abc}{bef} -> \frac{abc}{bef}
+    #  \fracabc -> \frac{a}{b}c
+    #  \sqrta -> \sqrt{a}
+    #  \sqrtab -> sqrt{a}b
+    final_answer = re.sub(r"(frac)([^{])(.)", "frac{\\2}{\\3}", final_answer)
+    final_answer = re.sub(r"(sqrt)([^{])", "sqrt{\\2}", final_answer)
+    final_answer = final_answer.replace("$", "")
+
+    # Normalize numbers
+    if final_answer.replace(",", "").isdigit():
+        final_answer = final_answer.replace(",", "")
+
+    return final_answer.strip()
+
+
+def is_correct_minerva(
+    solution_str: str, gt: str, answer_pattern: str = r"(?i)Answer\s*:\s*([^\n]+)"
+) -> tuple[bool, str]:
+    """Check if the solution is correct according to Minerva criteria.
+
+    Args:
+        solution_str: The solution string to check
+        gt: The ground truth answer
+        answer_pattern: Regex pattern to extract the answer
+
+    Returns:
+        is_correct: Whether the answer is correct
+    """
+    # Extract answer from solution
+    match = re.findall(answer_pattern, solution_str)
+    extracted_answer = match[-1] if match else "[INVALID]"
+    pred = normalize_final_answer(extracted_answer)
+    gt = normalize_final_answer(gt)
+
+    return pred == gt
 
 
 def accuracy_reward(response: str, ground_truth: str) -> float:
-    answer = extract_boxed_content(response)
-    return 1.0 if grade_answer(answer, ground_truth) else -1.0
+    return 1.0 if is_correct_minerva(response, ground_truth) else -1.0
 
 
 def soft_overlong_punishment(response_length: int, max_response_length: int, overlong_buffer_length: int):