Add Coaching & ScoreBoard class (result tracking) (#72)

* feat: Add Coach and ScoreBoard classes for performance tracking and difficulty adjustment
* feat: Add GroupedScores class to wrap aggregated scores
* refactor: Create ScoreStats class with tuple-based score statistics
* feat: Add unit test for Coach with CompositeDataset and multiple datasets
* fix: Add difficulty metadata to leg counting dataset
* feat: Add clear() method to ScoreBoard to reset all stored data
* feat: Add __len__ method to ScoreBoard to return number of scores
* feat: Add update_dataset_config method to CompositeDataset
* cleanup __init__ & imports

reasoning_gym/arithmetic/__init__.py

@@ -1,10 +1,5 @@
 """
 Arithmetic tasks for training reasoning capabilities:
-- Basic arithmetic
-- Chain sums
-- Word problems
-- Leg counting
-- Time intervals
 """
 
 from .basic_arithmetic import BasicArithmeticDataset, BasicArithmeticDatasetConfig
@@ -21,13 +16,10 @@ from .time_intervals import TimeIntervalsConfig, TimeIntervalsDataset
 __all__ = [
     "BasicArithmeticDataset",
     "BasicArithmeticDatasetConfig",
-    "basic_arithmetic_dataset",
     "ChainSum",
     "ChainSumConfig",
     "CalendarArithmeticConfig",
     "CalendarArithmeticDataset",
-    "Weekday",
-    "CalendarTask",
     "FractionSimplificationConfig",
     "FractionSimplificationDataset",
     "GCDConfig",

reasoning_gym/arithmetic/calendar_arithmetic.py

@@ -3,7 +3,7 @@ import math
 import random
 from dataclasses import dataclass
 from datetime import date, timedelta
-from enum import Enum, auto
+from enum import Enum, StrEnum, auto
 from typing import Any, Dict, List, Optional, Tuple
 
 from ..factory import ProceduralDataset, register_dataset
@@ -38,7 +38,7 @@ class Weekday(Enum):
         return self.name.capitalize()
 
 
-class CalendarTask(Enum):
+class CalendarTask(StrEnum):
     WEEKDAY_OFFSET = "weekday_offset"
     WEEKDAY_OF_DATE = "weekday_of_date"
     WEEKDAY_OF_DATE_FROM_FIRST_DATE = "weekday_of_date_from_first_day"

reasoning_gym/arithmetic/chain_sum.py

@@ -65,8 +65,10 @@ class ChainSum(ProceduralDataset):
             "question": f"{expression} =",
             "answer": str(result),
             "metadata": {
-                "num_terms": num_terms,
-                "num_digits": num_digits,
+                "difficulty": {
+                    "num_terms": num_terms,
+                    "num_digits": num_digits,
+                },
                 "expression": expression,
             },
         }

reasoning_gym/arithmetic/leg_counting.py

@@ -111,7 +111,13 @@ class LegCountingDataset(ProceduralDataset):
         return {
             "question": question,
             "answer": str(total_legs),
-            "metadata": {"animals": animals, "total_legs": total_legs},
+            "metadata": {
+                "difficulty": {
+                    "num_animals": len(animals),
+                },
+                "animals": animals,
+                "total_legs": total_legs,
+            },
         }
 
 

reasoning_gym/arithmetic/prime_factorization.py

@@ -2,7 +2,7 @@
 
 from dataclasses import dataclass
 from random import Random
-from typing import List, Optional, Tuple
+from typing import List, Optional
 
 from ..factory import ProceduralDataset, register_dataset