Hugging Face's logo

Datasets:
nmd2k
/
testcm

Modalities:
Text
Formats:
parquet
Size:
10K - 100K
Libraries:
Datasets
pandas
Dataset card Viewer Files Community
7
nmd2k commited on
Commit
f406645
·
verified ·
1 Parent(s): 441e8de

Upload folder using huggingface_hub

Browse files
Files changed (2) hide show
  1. README.md +129 -3
  2. codemmlu.py +12 -11
README.md CHANGED
@@ -1,3 +1,129 @@
1
- ---
2
- license: mit
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ dataset_info:
3
+ - config_name: api_frameworks
4
+ features:
5
+ - name: task_id
6
+ dtype: string
7
+ - name: question
8
+ dtype: string
9
+ - name: choices
10
+ sequence: string
11
+ splits:
12
+ - name: test
13
+ num_bytes: 12373682
14
+ num_examples: 19913
15
+ download_size: 2570062
16
+ dataset_size: 12373682
17
+ - config_name: code_completion
18
+ features:
19
+ - name: task_id
20
+ dtype: string
21
+ - name: question
22
+ dtype: string
23
+ - name: choices
24
+ sequence: string
25
+ splits:
26
+ - name: test
27
+ num_bytes: 12373682
28
+ num_examples: 19913
29
+ download_size: 2570062
30
+ dataset_size: 12373682
31
+ - config_name: code_repair
32
+ features:
33
+ - name: task_id
34
+ dtype: string
35
+ - name: question
36
+ dtype: string
37
+ - name: choices
38
+ sequence: string
39
+ splits:
40
+ - name: test
41
+ num_bytes: 12373682
42
+ num_examples: 19913
43
+ download_size: 2570062
44
+ dataset_size: 12373682
45
+ - config_name: dbms_sql
46
+ features:
47
+ - name: task_id
48
+ dtype: string
49
+ - name: question
50
+ dtype: string
51
+ - name: choices
52
+ sequence: string
53
+ splits:
54
+ - name: test
55
+ num_bytes: 12373682
56
+ num_examples: 19913
57
+ download_size: 2570062
58
+ dataset_size: 12373682
59
+ - config_name: defect_detection
60
+ features:
61
+ - name: task_id
62
+ dtype: string
63
+ - name: question
64
+ dtype: string
65
+ - name: choices
66
+ sequence: string
67
+ splits:
68
+ - name: test
69
+ num_bytes: 12373682
70
+ num_examples: 19913
71
+ download_size: 2570062
72
+ dataset_size: 12373682
73
+ - config_name: fill_in_the_middle
74
+ features:
75
+ - name: task_id
76
+ dtype: string
77
+ - name: question
78
+ dtype: string
79
+ - name: choices
80
+ sequence: string
81
+ splits:
82
+ - name: test
83
+ num_bytes: 12373682
84
+ num_examples: 19913
85
+ download_size: 2570062
86
+ dataset_size: 12373682
87
+ - config_name: others
88
+ features:
89
+ - name: task_id
90
+ dtype: string
91
+ - name: question
92
+ dtype: string
93
+ - name: choices
94
+ sequence: string
95
+ splits:
96
+ - name: test
97
+ num_bytes: 12373682
98
+ num_examples: 19913
99
+ download_size: 2570062
100
+ dataset_size: 12373682
101
+ - config_name: programming_syntax
102
+ features:
103
+ - name: task_id
104
+ dtype: string
105
+ - name: question
106
+ dtype: string
107
+ - name: choices
108
+ sequence: string
109
+ splits:
110
+ - name: test
111
+ num_bytes: 12373682
112
+ num_examples: 19913
113
+ download_size: 2570062
114
+ dataset_size: 12373682
115
+ - config_name: software_principles
116
+ features:
117
+ - name: task_id
118
+ dtype: string
119
+ - name: question
120
+ dtype: string
121
+ - name: choices
122
+ sequence: string
123
+ splits:
124
+ - name: test
125
+ num_bytes: 12373682
126
+ num_examples: 19913
127
+ download_size: 2570062
128
+ dataset_size: 12373682
129
+ ---
codemmlu.py CHANGED
@@ -88,17 +88,18 @@ class CodeMMLU(datasets.GeneratorBasedBuilder):
88
  def _generate_examples(self, iter_archive):
89
  """This function returns the examples in the raw (text) form."""
90
  for path, file in iter_archive:
91
- lines = (line.decode("utf-8") for line in file)
92
- reader = [json.loads(line) for line in lines]
93
- for data in reader:
94
- id_ = data['task_id']
 
95
 
96
- return_dict = {
97
- "question": data['question'],
98
- "choices": data['choices'],
99
- }
100
 
101
- if "fill_in_the_middle" in path:
102
- return_dict['problem_description'] = data['problem_description']
103
 
104
- yield id_, return_dict
 
88
  def _generate_examples(self, iter_archive):
89
  """This function returns the examples in the raw (text) form."""
90
  for path, file in iter_archive:
91
+ if path.endswith(".jsonl"):
92
+ lines = (line.decode("utf-8") for line in file)
93
+ reader = [json.loads(line) for line in lines]
94
+ for data in reader:
95
+ id_ = data['task_id']
96
 
97
+ return_dict = {
98
+ "question": data['question'],
99
+ "choices": data['choices'],
100
+ }
101
 
102
+ if "fill_in_the_middle" in path:
103
+ return_dict['problem_description'] = data['problem_description']
104
 
105
+ yield id_, return_dict