more improvements to mathquiz

configs/mathquiz.yaml

@@ -2,10 +2,7 @@ manager:
   args:
     max_rounds: 10
     win_conditions:
-      - TeacherCorrect
-      - TeacherIncorrect
       - StudentCorrect
-      - StudentIncorrect
 game:
   name: mathquiz
   players:

zero_sum_eval/game_manager.py

@@ -52,8 +52,8 @@ def do_eval(self, game_state):
                 break
             game_status = game_state.query_game()
             player = self.players[game_status.roles[0]]
-            logger.info(f"{player.id} turn {turn_count}:\n{game_state.display()}")
             game_state = self.do_turn(game_status, player)
+            logger.info(f"{player.id} turn {turn_count}:\n{game_state.display()}")
             round_count += 1
         return game_state
 

zero_sum_eval/games/mathquiz/mathquiz_game.py

@@ -66,29 +66,27 @@ def query_game(self):
         return self.initialize(
             environment=self.environment,
             context=new_context,
-            roles=new_roles
+            roles=new_roles,
+            target=self.target
         )
 
     def verify_answer(self, answer):
-        return str(answer) == str(self.target.strip())
+        return str(answer) == str(self.target)
 
     def validate_game(self):
         current_role = self.roles[0]
         if current_role == "TeacherGenerateQuestion":
             return None
-        if self.environment['teacher_answer'] is not None:
-            if current_role == "TeacherAnswerQuestion":
-                if self.verify_answer(self.environment['teacher_answer']):
-                    return "TeacherCorrect"
-                else:
-                    return "TeacherIncorrect"
-            elif current_role == "StudentAnswerQuestion":
-                if self.verify_answer(self.environment['teacher_answer']):
-                    return "StudentCorrect"
-                else:
-                    return "StudentIncorrect"
-        else:
-            return self.context['message']
+        elif current_role == "TeacherAnswerQuestion":
+            if self.verify_answer(self.environment['teacher_answer']):
+                return None
+            else:
+                return "TeacherIncorrect"
+        elif current_role == "StudentAnswerQuestion":
+            if self.verify_answer(self.environment['teacher_answer']):
+                return "StudentCorrect"
+            else:
+                return "StudentIncorrect"
 
     def get_next_roles(self, environment):
         if environment['question'] is None:

zero_sum_eval/games/mathquiz/mathquiz_player.py

@@ -10,25 +10,25 @@
 from zero_sum_eval.registry import PLAYER_REGISTRY, LM_REGISTRY
 
 class GenerateQuestion(dspy.Signature):
-    """Given a target answer, generate a challenging math question that has the target answer"""
+    """Given a target number, generate a challenging math question with the target number as the answer. Make sure not to include the answer in the question."""
 
-    target_answer = dspy.InputField(desc="target answer for the question")
-    math_question = dspy.OutputField(desc="math question with the target answer")
+    target_number = dspy.InputField(desc="target number")
+    math_question = dspy.OutputField(desc="math question with the target number as the answer")
 
 class AnswerQuestion(dspy.Signature):
-    """Given a math question, answer the question"""
+    """Given a math question, give the answer to the question as a number only"""
 
-    math_question = dspy.InputField(desc="math question to answer")
-    answer = dspy.OutputField(desc="answer to the math question")
+    math_question = dspy.InputField(desc="math question")
+    answer = dspy.OutputField(desc="answer to the math question with number only")
 
 class GenerateQuestionCoT(dspy.Module):
     def __init__(self):
         super().__init__()
         self.cot_question = dspy.ChainOfThought(GenerateQuestion)
 
 
-    def forward(self, target_answer):
-        cot_out = self.cot_question(target_answer=target_answer)
+    def forward(self, target_number):
+        cot_out = self.cot_question(target_number=target_number)
         return cot_out
 
 class AnswerQuestionCoT(dspy.Module):