mvasiliniuc commited on
Commit
b53a39b
·
1 Parent(s): 38d0bfb

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +177 -0
README.md ADDED
@@ -0,0 +1,177 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ annotations_creators:
3
+ - crowdsourced
4
+ license: other
5
+ language_creators:
6
+ - crowdsourced
7
+ language:
8
+ - code
9
+ task_categories:
10
+ - text-generation
11
+ tags:
12
+ - code, kotlin, native Android development
13
+ size_categories:
14
+ - 100K<n<1M
15
+ source_datasets: []
16
+ pretty_name: iva-kotlin-codeint-raw
17
+ task_ids:
18
+ - language-modeling
19
+ ---
20
+
21
+ # IVA Kotlin GitHub Code Dataset
22
+
23
+ ## Dataset Description
24
+
25
+ This is the raw IVA Kotlin dataset extracted from GitHub.
26
+ It contains uncurated Kotlin files gathered with the purpose to train a code generation model.
27
+
28
+ The dataset consists of 464215 kotlin code files from GitHub totaling ~361 MB of data.
29
+ The dataset was created from the public GitHub dataset on Google BiqQuery.
30
+
31
+ ### How to use it
32
+
33
+ To download the full dataset:
34
+
35
+ ```python
36
+ from datasets import load_dataset
37
+ dataset = load_dataset('mvasiliniuc/iva-kotlin-codeint', split='train')
38
+ ```
39
+
40
+ ```python
41
+
42
+ from datasets import load_dataset
43
+ dataset = load_dataset('mvasiliniuc/iva-kotlin-codeint', split='train')
44
+ print(dataset[723])
45
+
46
+ #OUTPUT:
47
+ {
48
+ "repo_name":"nemerosa/ontrack",
49
+ "path":"ontrack-extension-notifications/src/main/java/net/nemerosa/ontrack/extension/notifications/webhooks/WebhookController.kt",
50
+ "copies":"1",
51
+ "size":"3248",
52
+ "content":"...@RestController\n@RequestMapping(\"/extension/notifications/webhook\")\nclass WebhookController(\n private val webhookAdminService: WebhookAdminService,\n private val webhookExecutionService: ",
53
+ "license":"mit"
54
+ }
55
+ ```
56
+
57
+ ## Data Structure
58
+
59
+ ### Data Fields
60
+
61
+ |Field|Type|Description|
62
+ |---|---|---|
63
+ |repo_name|string|name of the GitHub repository|
64
+ |path|string|path of the file in GitHub repository|
65
+ |copies|string|number of occurrences in dataset|
66
+ |code|string|content of source file|
67
+ |size|string|size of the source file in bytes|
68
+ |license|string|license of GitHub repository|
69
+
70
+ ### Instance
71
+
72
+ ```json
73
+ {
74
+ "repo_name":"nemerosa/ontrack",
75
+ "path":"ontrack-extension-notifications/src/main/java/net/nemerosa/ontrack/extension/notifications/webhooks/WebhookController.kt",
76
+ "copies":"1",
77
+ "size":"3248",
78
+ "content":"...@RestController\n@RequestMapping(\"/extension/notifications/webhook\")\nclass WebhookController(\n private val webhookAdminService: WebhookAdminService,\n private val webhookExecutionService: ",
79
+ "license":"mit"
80
+ }
81
+ ```
82
+
83
+ ## Languages
84
+
85
+ The dataset contains only Kotlin files.
86
+
87
+ ```json
88
+ {
89
+ "Kotlin": [".kt"]
90
+ }
91
+ ```
92
+
93
+ ## Licenses
94
+
95
+ Each entry in the dataset contains the associated license. The following is a list of licenses involved and their occurrences.
96
+
97
+ ```json
98
+ {
99
+ "agpl-3.0": 9146,
100
+ "apache-2.0": 272388,
101
+ "artistic-2.0": 219,
102
+ "bsd-2-clause": 896,
103
+ "bsd-3-clause": 12328,
104
+ "cc0-1.0": 411,
105
+ "epl-1.0": 2111,
106
+ "gpl-2.0": 11080,
107
+ "gpl-3.0": 48911,
108
+ "isc": 997,
109
+ "lgpl-2.1": 297,
110
+ "lgpl-3.0": 7749,
111
+ "mit": 92540,
112
+ "mpl-2.0": 3386,
113
+ "unlicense": 1756
114
+ }
115
+ ```
116
+
117
+ ## Dataset Statistics
118
+
119
+ ```json
120
+ {
121
+ "Total size": "~361 MB",
122
+ "Number of files": 464215,
123
+ "Number of files under 500 bytes": 99845,
124
+ "Average file size in bytes": 3252,
125
+ }
126
+ ```
127
+
128
+ ## Dataset Creation
129
+
130
+ The dataset was created using Google Query for Github:
131
+ https://cloud.google.com/blog/topics/public-datasets/github-on-bigquery-analyze-all-the-open-source-code
132
+
133
+ The following steps were pursued for data
134
+ gathering:
135
+ 1. Creation of a dataset and a table in Google Big Query Project.
136
+ 2. Creation of a bucket in Google Cloud Storage.
137
+ 3. Creation of a query in Google Big Query Project.
138
+ 4. Running the query with the setting to output the results in the dataset and table
139
+ created at step one.
140
+ 5. Exporting the resulting dataset into the bucket created in step 2. Export format of JSON with gzip compression.
141
+
142
+ The result of these steps leads to the following results:
143
+ * 2.7 TB Processed,
144
+ * number of extracted rows/files was 464,215
145
+ * total logical bytes 1.46 GB.
146
+ * the result amounts to 7 json.gz files in a total of 361 MB.
147
+
148
+ The SQL Query used is:
149
+ ```sql
150
+ SELECT
151
+ f.repo_name, f.path, c.copies, c.size, c.content, l.license
152
+ FROM
153
+ (select f.*, row_number() over (partition by id order by path desc) as seqnum from `bigquery-public-data.github_repos.files` AS f) f
154
+ JOIN
155
+ `bigquery-public-data.github_repos.contents` AS c
156
+ ON
157
+ f.id = c.id AND seqnum=1
158
+ JOIN
159
+ `bigquery-public-data.github_repos.licenses` AS l
160
+ ON
161
+ f.repo_name = l.repo_name
162
+ WHERE
163
+ NOT c.binary AND ((f.path LIKE '%.kt') AND (c.size BETWEEN 0 AND 1048575))
164
+ ```
165
+
166
+ ## Data Splits
167
+ The dataset only contains a train split.
168
+
169
+ Using the curated version of this dataset, a split was made into multiple repositories:
170
+ * Clean Version: https://huggingface.co/datasets/mvasiliniuc/iva-kotlin-codeint-clean
171
+ * Clean Version Train: https://huggingface.co/datasets/mvasiliniuc/iva-kotlin-codeint-clean-train
172
+ * Clean Version Valid: https://huggingface.co/datasets/mvasiliniuc/iva-kotlin-codeint-clean-valid
173
+
174
+ # Considerations for Using the Data
175
+
176
+ The dataset consists of source code from a wide range of repositories.
177
+ As such they can potentially include harmful or biased code as well as sensitive information like passwords or usernames.