dm_math.py 6.84 KB
Newer Older
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
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
"""
Analysing Mathematical Reasoning Abilities of Neural Models
https://arxiv.org/pdf/1904.01557.pdf

*Describe dataset here*

Homepage: https://github.com/deepmind/mathematics_dataset
"""
from lm_eval.metrics import mean
from lm_eval.base import Task, rf


_CITATION = """
ADD CITATION HERE
"""


class DMMath(Task):
    DATASET_PATH = "math_dataset"
    DATASET_NAME = None

    def has_training_docs(self):
        return True

    def has_validation_docs(self):
        return False

    def has_test_docs(self):
        return True

    def training_docs(self):
        return map(self._process_doc, self.dataset["train"])

    def validation_docs(self):
        return NotImplemented

    def test_docs(self):
        return map(self._process_doc, self.dataset["test"])

    def _process_doc(self, doc):
        # dataset fields are in format "b'{string contents here}'". 
        # we want to strip this extraneous bytes formatting from the strings.
        doc["answer"] = doc["answer"].lstrip("b'").rstrip("'")
        doc["question"] = doc["question"].lstrip("b'").rstrip("'")
        return doc

    def doc_to_text(self, doc):
        return doc["question"] + "Answer:"

    def should_decontaminate(self):
        return True

    def doc_to_decontamination_query(self, doc):
        return doc["question"]

    def doc_to_target(self, doc):
        return " " + doc["answer"]

    def construct_requests(self, doc, ctx):
        return rf.greedy_until(ctx, ["\n", "\n\n"])

    def process_results(self, doc, results):
        if doc["answer"].rstrip("\n") == results[0].rstrip("\n"): # for now, simple string comparison. TODO: sympy answer checking, especially for harder subsets that don't just return a number
            is_correct = 1
        else:
            is_correct = 0
        return {"acc": is_correct}

    def aggregation(self):
        return {"acc": mean}

    def higher_is_better(self):
        return {"acc": True}


class DMMathLinAlg1d(DMMath):
    VERSION = 0
    DATASET_NAME = "algebra__linear_1d"

class DMMathLinAlg1dComp(DMMath):
    VERSION = 0
    DATASET_NAME = "algebra__linear_1d_composed"

class DMMathLinAlg2d(DMMath):
    VERSION = 0
    DATASET_NAME = "algebra__linear_2d"

class DMMathLinAlg2dComp(DMMath):
    VERSION = 0
    DATASET_NAME = "algebra__linear_2d"

class DMMathPolyRoots(DMMath):
    VERSION = 0
    DATASET_NAME = "algebra__polynomial_roots"

class DMMathPolyRootsComp(DMMath):
    VERSION = 0
    DATASET_NAME = "algebra__polynomial_roots_composed"

class DMMathSeqNext(DMMath):
    VERSION = 0
    DATASET_NAME = "algebra__sequence_next_term"

class DMMathSeqNth(DMMath):
    VERSION = 0
    DATASET_NAME = "algebra__sequence_nth_term"

class DMMathAddOrSub(DMMath):
    VERSION = 0
    DATASET_NAME = "arithmetic__add_or_sub"

class DMMathAddOrSubBase(DMMath):
    VERSION = 0
    DATASET_NAME = "arithmetic__add_or_sub_in_base"

class DMMathAddOrSubComp(DMMath):
    VERSION = 0
    DATASET_NAME = "arithmetic__add_sub_multiple"

class DMMathDiv(DMMath):
    VERSION = 0
    DATASET_NAME = "arithmetic__div"

class DMMathMixed(DMMath):
    VERSION = 0
    DATASET_NAME = "arithmetic__mixed"

class DMMathMult(DMMath):
    VERSION = 0
    DATASET_NAME = "arithmetic__mul"

class DMMathMultDivComp(DMMath):
    VERSION = 0
    DATASET_NAME = "arithmetic__mul_div_multiple"

class DMMathNearestRoot(DMMath):
    VERSION = 0
    DATASET_NAME = "arithmetic__nearest_integer_root"

class DMMathSimplify(DMMath):
    VERSION = 0
    DATASET_NAME = "arithmetic__simplify_surd"

class DMMathDiff(DMMath):
    VERSION = 0
    DATASET_NAME = "calculus__differentiate"

class DMMathDiffComp(DMMath):
    VERSION = 0
    DATASET_NAME = "calculus__differentiate_composed"

class DMMathClosest(DMMath):
    VERSION = 0
    DATASET_NAME = "comparison__closest"

class DMMathClosestComp(DMMath):
    VERSION = 0
    DATASET_NAME = "comparison__closest_composed"

class DMMathKthBiggest(DMMath):
    VERSION = 0
    DATASET_NAME = "comparison__kth_biggest"

class DMMathKthBiggestComp(DMMath):
    VERSION = 0
    DATASET_NAME = "comparison__kth_biggest_composed"

class DMMathPair(DMMath):
    VERSION = 0
    DATASET_NAME = "comparison__pair"

class DMMathPairComp(DMMath):
    VERSION = 0
    DATASET_NAME = "comparison__pair_composed"

class DMMathSort(DMMath):
    VERSION = 0
    DATASET_NAME = "comparison__sort"

class DMMathSortComp(DMMath):
    VERSION = 0
    DATASET_NAME = "comparison__sort_composed"

class DMMathMeasConv(DMMath):
    VERSION = 0
    DATASET_NAME = "measurement__conversion"

class DMMathMeasTime(DMMath):
    VERSION = 0
    DATASET_NAME = "measurement__time"

class DMMathBaseConv(DMMath):
    VERSION = 0
    DATASET_NAME = "numbers__base_conversion"

class DMMathDivRemainder(DMMath):
    VERSION = 0
    DATASET_NAME = "numbers__div_remainder"

class DMMathDivRemainderComp(DMMath):
    VERSION = 0
    DATASET_NAME = "numbers__div_remainder_composed"

class DMMathGcd(DMMath):
    VERSION = 0
    DATASET_NAME = "numbers__gcd"

class DMMathGcdComp(DMMath):
    VERSION = 0
    DATASET_NAME = "numbers__gcd_composed"

class DMMathIsFactor(DMMath):
    VERSION = 0
    DATASET_NAME = "numbers__is_factor"

class DMMathIsFactorComp(DMMath):
    VERSION = 0
    DATASET_NAME = "numbers__is_factor_composed"

class DMMathLcm(DMMath):
    VERSION = 0
    DATASET_NAME = "numbers__lcm"

class DMMathLcmComp(DMMath):
    VERSION = 0
    DATASET_NAME = "numbers__lcm_composed"

class DMMathListPrimeFactors(DMMath):
    VERSION = 0
    DATASET_NAME = "numbers__list_prime_factors"

class DMMathListPrimeFactorsComp(DMMath):
    VERSION = 0
    DATASET_NAME = "numbers__list_prime_factors_composed"

class DMMathPlaceVal(DMMath):
    VERSION = 0
    DATASET_NAME = "numbers__place_value"

class DMMathPlaceValComp(DMMath):
    VERSION = 0
    DATASET_NAME = "numbers__place_value_composed"

class DMMathRoundNum(DMMath):
    VERSION = 0
    DATASET_NAME = "numbers__round_number"

class DMMathRoundNumComp(DMMath):
    VERSION = 0
    DATASET_NAME = "numbers__round_number_composed"

class DMMathAddPoly(DMMath):
    VERSION = 0
    DATASET_NAME = "polynomials__add"

class DMMathPolyCoeff(DMMath):
    VERSION = 0
    DATASET_NAME = "polynomials__coefficient_named"

class DMMathPolyCollect(DMMath):
    VERSION = 0
    DATASET_NAME = "polynomials__collect"

class DMMathPolyComp(DMMath):
    VERSION = 0
    DATASET_NAME = "polynomials__compose"

class DMMathPolyEval(DMMath):
    VERSION = 0
    DATASET_NAME = "polynomials__evaluate"

class DMMathPolyEvalComp(DMMath):
    VERSION = 0
    DATASET_NAME = "polynomials__evaluate_composed"

class DMMathPolyExpand(DMMath):
    VERSION = 0
    DATASET_NAME = "polynomials__expand"

class DMMathPolySimplify(DMMath):
    VERSION = 0
    DATASET_NAME = "polynomials__simplify_power"

class DMMathProbLevelSet(DMMath):
    VERSION = 0
    DATASET_NAME = "probability__swr_p_level_set"

class DMMathProbSeq(DMMath):
    VERSION = 0
    DATASET_NAME = "probability__swr_p_sequence"