open-thought/reasoning-gym
1
0
Fork 0
mirror of https://github.com/open-thought/reasoning-gym.git synced 2026-04-29 17:35:16 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

jugs jugs jugs lint

Browse source
This commit is contained in:
Rich Jones 2025-02-20 16:15:29 +01:00
parent 6f00690ae1
commit 0f798457ed
2 changed files with 32 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

30
tests/test_jugs.py
View file

@ -7,7 +7,7 @@ from reasoning_gym.algorithmic.jugs import JugsConfig, JugsDataset
def test_jugs():
"""Test basic properties and solution of generated items"""
config = JugsConfig(seed=42, size=10, num_jugs=3, difficulty=20)
config = JugsConfig(seed=42, size=1000, num_jugs=3, difficulty=5)
dataset = JugsDataset(config)
# easy
@ -20,3 +20,31 @@ def test_jugs():
# Test the scoring
assert dataset.score_answer(answer=json.dumps(item["metadata"]["possible_answer"]), entry=item) == 1.0
assert dataset.score_answer(answer=None, entry=item) == 0.0
config = JugsConfig(seed=42, size=1, num_jugs=3, difficulty=50)
dataset = JugsDataset(config)
# med
for item in dataset:
assert isinstance(item, dict)
assert "question" in item
assert "answer" in item
assert "metadata" in item
# Test the scoring
assert dataset.score_answer(answer=json.dumps(item["metadata"]["possible_answer"]), entry=item) == 1.0
assert dataset.score_answer(answer=None, entry=item) == 0.0
config = JugsConfig(seed=42, size=1, num_jugs=3, difficulty=99)
dataset = JugsDataset(config)
# hard
for item in dataset:
assert isinstance(item, dict)
assert "question" in item
assert "answer" in item
assert "metadata" in item
# Test the scoring
assert dataset.score_answer(answer=json.dumps(item["metadata"]["possible_answer"]), entry=item) == 1.0
assert dataset.score_answer(answer=None, entry=item) == 0.0