mirror of
https://github.com/open-thought/reasoning-gym.git
synced 2026-04-19 12:58:07 +00:00
98 lines
3.3 KiB
Python
98 lines
3.3 KiB
Python
"""Letter counting task generator"""
|
|
|
|
import re
|
|
from dataclasses import dataclass
|
|
from random import Random
|
|
from typing import Optional
|
|
|
|
from reasoning_gym.data import read_data_file
|
|
|
|
from ..coaching import BaseCurriculum, RangeAttributeDefinition
|
|
from ..factory import ProceduralDataset, register_dataset
|
|
|
|
DATASET_NAME = "letter_counting"
|
|
|
|
|
|
@dataclass
|
|
class LetterCountingConfig:
|
|
"""Configuration for letter counting task generation"""
|
|
|
|
min_words: int = 5 # Minimum words in span
|
|
max_words: int = 15 # Maximum words in span
|
|
seed: Optional[int] = None
|
|
size: int = 500 # Virtual dataset size
|
|
|
|
def validate(self) -> None:
|
|
"""Validate configuration parameters"""
|
|
assert self.min_words > 0, "min_words must be positive"
|
|
assert self.max_words >= self.min_words, "max_words must be >= min_words"
|
|
|
|
|
|
class LetterCountingDataset(ProceduralDataset):
|
|
"""Generates letter counting tasks from text spans"""
|
|
|
|
def __init__(self, config: LetterCountingConfig):
|
|
super().__init__(config=config, seed=config.seed, size=config.size)
|
|
|
|
# Load and preprocess text
|
|
text = read_data_file("in_the_year_2889.txt")
|
|
# Extract words and clean them to contain only alphanumeric characters
|
|
self.words = [word for word in re.findall(r"\b\w+\b", text) if word.isalnum()]
|
|
|
|
def __getitem__(self, idx: int) -> dict:
|
|
"""Generate a single letter counting task"""
|
|
rng = Random(self.seed + idx)
|
|
|
|
# Select random span of words
|
|
span_length = min(
|
|
rng.randint(self.config.min_words, self.config.max_words),
|
|
len(self.words),
|
|
)
|
|
start_idx = rng.randint(0, len(self.words) - span_length)
|
|
span = self.words[start_idx : start_idx + span_length]
|
|
|
|
# Get all unique letters from span
|
|
letters = set("".join(span).lower())
|
|
if not letters:
|
|
letters = {"a"} # Fallback if span has no letters
|
|
|
|
# Select random letter that appears in the span
|
|
target_letter = rng.choice(sorted(letters))
|
|
|
|
# Count occurrences
|
|
count = sum(word.lower().count(target_letter) for word in span)
|
|
|
|
return {
|
|
"question": f'How many times does the letter "{target_letter}" appear in the text: "{" ".join(span)}"?',
|
|
"answer": str(count),
|
|
"metadata": {
|
|
"source_dataset": DATASET_NAME,
|
|
"source_index": idx,
|
|
"span_length": span_length,
|
|
"target_letter": target_letter,
|
|
"span": span,
|
|
"difficulty": {
|
|
"words": (self.config.min_words, self.config.max_words),
|
|
},
|
|
},
|
|
}
|
|
|
|
|
|
class LetterCountingCurriculum(BaseCurriculum):
|
|
def __init__(self):
|
|
super().__init__(LetterCountingCurriculum.__name__, LetterCountingConfig)
|
|
|
|
# Define attributes
|
|
self._define_attributes(
|
|
RangeAttributeDefinition(
|
|
name="words",
|
|
levels=list(range(5, 20, 2)),
|
|
description="Number of words in the span",
|
|
lower_field_name="min_words",
|
|
upper_field_name="max_words",
|
|
ensure_interval=True,
|
|
),
|
|
)
|
|
|
|
|
|
register_dataset(DATASET_NAME, LetterCountingDataset, LetterCountingConfig, LetterCountingCurriculum)
|