File size: 5,847 Bytes
a2c2f1e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
"""Frontier-CS: Evolving Challenges for Evolving Intelligence"""

import os
import datasets

_DESCRIPTION = """
Frontier-CS is an unsolved, open-ended, verifiable, and diverse benchmark
for evaluating AI on challenging computer science problems.
"""

_HOMEPAGE = "https://frontier-cs.org"
_LICENSE = "Apache-2.0"

class FrontierCSConfig(datasets.BuilderConfig):
    def __init__(self, **kwargs):
        super().__init__(**kwargs)

class FrontierCS(datasets.GeneratorBasedBuilder):
    VERSION = datasets.Version("1.0.0")

    BUILDER_CONFIGS = [
        FrontierCSConfig(name="algorithmic", description="118 algorithmic problems"),
        FrontierCSConfig(name="research", description="63 research problems"),
        FrontierCSConfig(name="all", description="All problems"),
    ]

    DEFAULT_CONFIG_NAME = "all"

    def _info(self):
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features({
                "problem_id": datasets.Value("string"),
                "category": datasets.Value("string"),
                "statement": datasets.Value("string"),
                "config": datasets.Value("string"),
            }),
            homepage=_HOMEPAGE,
            license=_LICENSE,
        )

    def _split_generators(self, dl_manager):
        return [
            datasets.SplitGenerator(
                name=datasets.Split.TEST,
                gen_kwargs={"config_name": self.config.name},
            ),
        ]

    def _generate_examples(self, config_name):
        base_path = os.path.dirname(os.path.abspath(__file__))
        idx = 0

        # Algorithmic problems
        if config_name in ["algorithmic", "all"]:
            algo_path = os.path.join(base_path, "algorithmic", "problems")
            if os.path.exists(algo_path):
                for problem_id in sorted(os.listdir(algo_path)):
                    problem_dir = os.path.join(algo_path, problem_id)
                    if not os.path.isdir(problem_dir):
                        continue

                    statement = ""
                    config = ""

                    statement_file = os.path.join(problem_dir, "statement.txt")
                    if os.path.exists(statement_file):
                        with open(statement_file, "r", encoding="utf-8") as f:
                            statement = f.read()

                    config_file = os.path.join(problem_dir, "config.yaml")
                    if os.path.exists(config_file):
                        with open(config_file, "r", encoding="utf-8") as f:
                            config = f.read()

                    yield idx, {
                        "problem_id": problem_id,
                        "category": "algorithmic",
                        "statement": statement,
                        "config": config,
                    }
                    idx += 1

        # Research problems
        if config_name in ["research", "all"]:
            research_path = os.path.join(base_path, "research", "problems")
            if os.path.exists(research_path):
                for category in sorted(os.listdir(research_path)):
                    category_dir = os.path.join(research_path, category)
                    if not os.path.isdir(category_dir):
                        continue

                    # Check if this is a problem dir or a parent dir
                    readme_file = os.path.join(category_dir, "readme")
                    config_file = os.path.join(category_dir, "config.yaml")

                    if os.path.exists(readme_file) or os.path.exists(config_file):
                        # This is a problem directory
                        statement = ""
                        config = ""

                        if os.path.exists(readme_file):
                            with open(readme_file, "r", encoding="utf-8") as f:
                                statement = f.read()

                        if os.path.exists(config_file):
                            with open(config_file, "r", encoding="utf-8") as f:
                                config = f.read()

                        yield idx, {
                            "problem_id": category,
                            "category": "research",
                            "statement": statement,
                            "config": config,
                        }
                        idx += 1
                    else:
                        # Check subdirectories
                        for subproblem in sorted(os.listdir(category_dir)):
                            subproblem_dir = os.path.join(category_dir, subproblem)
                            if not os.path.isdir(subproblem_dir):
                                continue

                            readme_file = os.path.join(subproblem_dir, "readme")
                            config_file = os.path.join(subproblem_dir, "config.yaml")

                            statement = ""
                            config = ""

                            if os.path.exists(readme_file):
                                with open(readme_file, "r", encoding="utf-8") as f:
                                    statement = f.read()

                            if os.path.exists(config_file):
                                with open(config_file, "r", encoding="utf-8") as f:
                                    config = f.read()

                            if statement or config:
                                yield idx, {
                                    "problem_id": f"{category}/{subproblem}",
                                    "category": "research",
                                    "statement": statement,
                                    "config": config,
                                }
                                idx += 1