Icannos commited on
Commit
fe38569
1 Parent(s): 368d66f

loading script

Browse files
chess_studies.py ADDED
@@ -0,0 +1,75 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ """Chess studies and annotated games from the top lichess studies and from https://www.angelfire.com/games3/smartbridge/"""
3
+
4
+
5
+ import os
6
+ import pandas as pd
7
+
8
+ import datasets
9
+
10
+
11
+ _CITATION = """TO COME."""
12
+
13
+ _DESCRIPTION = """\
14
+ Chess studies and annotated games from the top lichess studies and from https://www.angelfire.com/games3/smartbridge/
15
+ This dataset consists of annotated chess games from several sources and aggregated into a single dataset. It is intended
16
+ to train language models to generate chess games and studies.
17
+ """
18
+
19
+ _HOMEPAGE = ""
20
+
21
+ _LICENSE = "CC0"
22
+
23
+ _URLS = {
24
+ "lichess": "lichess_studies.csv",
25
+ "others": "others.csv",
26
+ }
27
+
28
+
29
+ class ChessStudies(datasets.GeneratorBasedBuilder):
30
+
31
+ VERSION = datasets.Version("1.0.0")
32
+
33
+
34
+ BUILDER_CONFIGS = [
35
+ datasets.BuilderConfig(name="lichess", version=VERSION, description="Top studies scrapped from the lichess website"),
36
+ datasets.BuilderConfig(name="others", version=VERSION, description="Studies aggregated from other sources"),
37
+ ]
38
+
39
+ DEFAULT_CONFIG_NAME = "lichess"
40
+
41
+ def _info(self):
42
+ features = datasets.Features(
43
+ {
44
+ "text": datasets.Value("string"),
45
+ }
46
+ )
47
+ return datasets.DatasetInfo(
48
+ description=_DESCRIPTION,
49
+ features=features,
50
+ homepage=_HOMEPAGE,
51
+ license=_LICENSE,
52
+ citation=_CITATION,
53
+ )
54
+
55
+ def _split_generators(self, dl_manager):
56
+ urls = _URLS[self.config.name]
57
+ data_dir = dl_manager.download(urls)
58
+ return [
59
+ datasets.SplitGenerator(
60
+ name=datasets.Split.TRAIN,
61
+ # These kwargs will be passed to _generate_examples
62
+ gen_kwargs={
63
+ "filepath": os.path.join(data_dir),
64
+ "split": "train",
65
+ },
66
+ ),
67
+ ]
68
+
69
+ # method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
70
+ def _generate_examples(self, filepath, split):
71
+ df = pd.read_csv(filepath)
72
+ for i, row in df.iterrows():
73
+ yield i, {
74
+ "text": row['text'],
75
+ }
train.csv → lichess_studies.csv RENAMED
File without changes
others.csv ADDED
The diff for this file is too large to render. See raw diff