Clémentine commited on
Commit
da97b7d
1 Parent(s): b91b2b0
induction/test.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
induction/train.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
induction/validation.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
pattern_match/test.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
pattern_match/train.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
pattern_match/validation.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
synthetic_reasoning.py ADDED
@@ -0,0 +1,61 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import datasets
2
+ import os
3
+ import json
4
+
5
+
6
+ _CITATION = """
7
+
8
+ """
9
+
10
+ _DESCRIPTION = """
11
+
12
+ """
13
+
14
+ class Builder(datasets.GeneratorBasedBuilder):
15
+ VERSION = datasets.Version("1.0.0")
16
+
17
+ BUILDER_CONFIGS = [
18
+ datasets.BuilderConfig(name=name, version=datasets.Version("1.0.0"), description=_DESCRIPTION)
19
+ for name in ["variable_substitution", "pattern_match", "induction"]
20
+ ]
21
+
22
+ def _info(self):
23
+ features = datasets.Features(
24
+ {
25
+ "source": datasets.Value("string"),
26
+ "target": datasets.Value("string"),
27
+ }
28
+ )
29
+ return datasets.DatasetInfo(
30
+ description=_DESCRIPTION,
31
+ features=features,
32
+ homepage="",
33
+ license="",
34
+ citation=_CITATION,
35
+ )
36
+
37
+ def _split_generators(self, dl_manager):
38
+ train_json = dl_manager.download(os.path.join(self.config.name, "train.jsonl"))
39
+ valid_json = dl_manager.download(os.path.join(self.config.name, "validation.jsonl"))
40
+ test_json = dl_manager.download(os.path.join(self.config.name, "test.jsonl"))
41
+
42
+ return [
43
+ datasets.SplitGenerator(
44
+ name=datasets.Split.TRAIN,
45
+ gen_kwargs={"path": train_json},
46
+ ),
47
+ datasets.SplitGenerator(
48
+ name=datasets.Split.TEST,
49
+ gen_kwargs={"path": test_json},
50
+ ),
51
+ datasets.SplitGenerator(
52
+ name=datasets.Split.VALIDATION,
53
+ gen_kwargs={"path": valid_json},
54
+ ),
55
+ ]
56
+
57
+ # method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
58
+ def _generate_examples(self, path):
59
+ with open(path, encoding="utf-8") as f:
60
+ for key, row in enumerate(f):
61
+ yield key, json.loads(row)
variable_substitution/test.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
variable_substitution/train.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
variable_substitution/validation.jsonl ADDED
The diff for this file is too large to render. See raw diff