OpenChristianData commited on
Commit
031bae2
·
verified ·
1 Parent(s): c2b4c0e

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +169 -173
README.md CHANGED
@@ -1,173 +1,169 @@
1
- ---
2
- license: cc0-1.0
3
- language:
4
- - en
5
- pretty_name: Open Christian Data
6
- size_categories:
7
- - 100K<n<1M
8
- task_categories:
9
- - text-generation
10
- - question-answering
11
- - text-classification
12
- - feature-extraction
13
- tags:
14
- - christian
15
- - theology
16
- - bible
17
- - commentary
18
- - sermon
19
- - catechism
20
- - confession
21
- - church-fathers
22
- - public-domain
23
- - religion
24
- configs:
25
- - config_name: bible_text
26
- data_files:
27
- - path: data/bible_text.jsonl
28
- split: train
29
- - config_name: catechism_qa
30
- data_files:
31
- - path: data/catechism_qa.jsonl
32
- split: train
33
- - config_name: church_fathers
34
- data_files:
35
- - path: data/church_fathers.jsonl
36
- split: train
37
- - config_name: commentary
38
- data_files:
39
- - path: data/commentary.jsonl
40
- split: train
41
- - config_name: devotional
42
- data_files:
43
- - path: data/devotional.jsonl
44
- split: train
45
- - config_name: doctrinal_document
46
- data_files:
47
- - path: data/doctrinal_document.jsonl
48
- split: train
49
- - config_name: hymn_collection
50
- data_files:
51
- - path: data/hymn_collection.jsonl
52
- split: train
53
- - config_name: prayer
54
- data_files:
55
- - path: data/prayer.jsonl
56
- split: train
57
- - config_name: reference_entry
58
- data_files:
59
- - path: data/reference_entry.jsonl
60
- split: train
61
- - config_name: sermon
62
- data_files:
63
- - path: data/sermon.jsonl
64
- split: train
65
- - config_name: structured_text
66
- data_files:
67
- - path: data/structured_text.jsonl
68
- split: train
69
- - config_name: topical_reference
70
- data_files:
71
- - path: data/topical_reference.jsonl
72
- split: train
73
- ---
74
-
75
- # Open Christian Data
76
-
77
- A free, author/work-led dataset of public domain Christian literature — Bible text, commentaries, sermons, catechisms, confessions, theology, prayer, and reference works, projected into clean records with full provenance metadata. Intentionally designed for developers and AI/LLM training.
78
-
79
- All content is written by authors who died before 1928. All texts are in the public domain. The dataset is CC0.
80
-
81
- ## Export shape
82
-
83
- The live whole-corpus HuggingFace export is the per-schema JSONL export generated by
84
- `build/scripts/export_huggingface.py`. The `original`/`modernised` export path in the source
85
- repository is a separate NSH-style publish-projection path and is not the current whole-corpus
86
- HuggingFace release.
87
-
88
- ## What's included
89
-
90
- The public inventory is counted by recognized work units: the title-level item a reader, library, citation, or source catalog would treat as present. Export records, verses, dictionary entries, topics, Q&A pairs, chapters, and quotations are subordinate technical units.
91
-
92
- Generated from current `data/` with `py -3 build/tools/count_dataset_records.py`.
93
-
94
- | Category | Work units | Top-level records |
95
- |---|---:|---:|
96
- | Bible text | 9 | 282,395 |
97
- | Catechisms | 15 | 3,509 |
98
- | Church Fathers quotation source | 1 | 70,164 |
99
- | Commentaries | 34 | 113,903 |
100
- | Devotionals | 2 | 1,464 |
101
- | Doctrinal documents | 38 | 38 |
102
- | Hymns | 1 | 34,904 |
103
- | Lexicon | 1 | 0 |
104
- | Prayers | 4 | 205 |
105
- | Reference works | 7 | 26,305 |
106
- | Sermons | 7 | 5,297 |
107
- | Structured texts | 288 | 289 |
108
- | Topical reference works | 1 | 5,322 |
109
- | **Total** | **408** | **543,795** |
110
-
111
- The live HuggingFace export flattens structured texts and doctrinal documents, so this release exports **803,985 JSONL rows**. The source repository includes the generated work metadata audit at `docs/WORK_CATALOG.md` and a browser-friendly review surface at `docs/WORK_CATALOG.html`.
112
-
113
- ## Loading the data
114
-
115
- ```python
116
- from datasets import load_dataset
117
-
118
- # Load a single config
119
- original = load_dataset("OpenChristianData/open-christian-data", "original")
120
- modernised = load_dataset("OpenChristianData/open-christian-data", "modernised")
121
-
122
- # Filter to a specific author
123
- wesley = [r for r in modernised["train"] if r["meta"]["author_slug"] == "wesley"]
124
-
125
- # Filter to works with no modernised sibling by reading the coverage table above
126
- ```
127
-
128
- ## Data structure
129
-
130
- Every record is a flattened row combining the resource-level metadata with entry-level fields. Key fields present on all records:
131
-
132
- | Field | Description |
133
- |---|---|
134
- | `_source_id` | Unique identifier for the parent resource (e.g., `matthew-henry-complete`) |
135
- | `entry_id` | Unique identifier for this entry (e.g., `matthew-henry-complete.Ezek.1.1-3`) |
136
- | `author` | Author name |
137
- | `tradition` | Array of theological tradition labels (e.g., `["reformed", "puritan"]`) |
138
- | `era` | Historical era (`patristic`, `medieval`, `reformation`, `post-reformation`, `modern`) |
139
- | `license` | `cc0-1.0` or `public-domain` |
140
- | `schema_type` | Content type — one of the config names above |
141
-
142
- Schema-specific fields vary by content type. Full field documentation is in the [source repository](https://github.com/OpenChristianData/open-christian-data/blob/main/docs/SCHEMA_SPEC.md).
143
-
144
- ## Provenance
145
-
146
- Every record carries full provenance — the source URL, source format, download date, processing script, and a SHA-256 hash of the source file. This is stored in the `provenance_*` fields in the flattened export.
147
-
148
- ## Verse references
149
-
150
- All verse references use [OSIS format](https://crosswire.org/osis/) (`Gen.1.1`, `Rom.9.1-Rom.9.5`). Where sources use non-standard reference formats, both the raw form and the normalized OSIS form are stored.
151
-
152
- ## License
153
-
154
- - **Data**: [CC0 1.0 Universal](https://creativecommons.org/publicdomain/zero/1.0/) — all underlying texts are public domain; the structuring and provenance tracking are dedicated to the public domain
155
- - **Bible text (BSB)**: [CC0 1.0](https://berean.bible/licensing.htm)
156
- - **Code and schemas**: [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) — free for non-commercial use; commercial use requires a license (see [source repository](https://github.com/OpenChristianData/open-christian-data))
157
-
158
- ## Citation
159
-
160
- ```bibtex
161
- @dataset{open_christian_data_2026,
162
- title = {Open Christian Data},
163
- author = {OpenChristianData},
164
- year = {2026},
165
- publisher = {HuggingFace},
166
- url = {https://huggingface.co/datasets/OpenChristianData/open-christian-data},
167
- license = {CC0-1.0}
168
- }
169
- ```
170
-
171
- ## Source repository
172
-
173
- [github.com/OpenChristianData/open-christian-data](https://github.com/OpenChristianData/open-christian-data)
 
1
+ ---
2
+ license: cc0-1.0
3
+ language:
4
+ - en
5
+ pretty_name: Open Christian Data
6
+ size_categories:
7
+ - 100K<n<1M
8
+ task_categories:
9
+ - text-generation
10
+ - question-answering
11
+ - text-classification
12
+ - feature-extraction
13
+ tags:
14
+ - christian
15
+ - theology
16
+ - bible
17
+ - commentary
18
+ - sermon
19
+ - catechism
20
+ - confession
21
+ - church-fathers
22
+ - public-domain
23
+ - religion
24
+ configs:
25
+ - config_name: bible_text
26
+ data_files:
27
+ - path: data/bible_text.jsonl
28
+ split: train
29
+ - config_name: catechism_qa
30
+ data_files:
31
+ - path: data/catechism_qa.jsonl
32
+ split: train
33
+ - config_name: church_fathers
34
+ data_files:
35
+ - path: data/church_fathers.jsonl
36
+ split: train
37
+ - config_name: commentary
38
+ data_files:
39
+ - path: data/commentary.jsonl
40
+ split: train
41
+ - config_name: devotional
42
+ data_files:
43
+ - path: data/devotional.jsonl
44
+ split: train
45
+ - config_name: doctrinal_document
46
+ data_files:
47
+ - path: data/doctrinal_document.jsonl
48
+ split: train
49
+ - config_name: hymn_collection
50
+ data_files:
51
+ - path: data/hymn_collection.jsonl
52
+ split: train
53
+ - config_name: prayer
54
+ data_files:
55
+ - path: data/prayer.jsonl
56
+ split: train
57
+ - config_name: reference_entry
58
+ data_files:
59
+ - path: data/reference_entry.jsonl
60
+ split: train
61
+ - config_name: sermon
62
+ data_files:
63
+ - path: data/sermon.jsonl
64
+ split: train
65
+ - config_name: structured_text
66
+ data_files:
67
+ - path: data/structured_text.jsonl
68
+ split: train
69
+ - config_name: topical_reference
70
+ data_files:
71
+ - path: data/topical_reference.jsonl
72
+ split: train
73
+ ---
74
+
75
+ # Open Christian Data
76
+
77
+ A free, author/work-led dataset of public domain Christian literature — Bible text, commentaries, sermons, catechisms, confessions, theology, prayer, and reference works, projected into clean records with full provenance metadata. Intentionally designed for developers and AI/LLM training.
78
+
79
+ Publication status: The authorized 2026-07-17 upload to [OpenChristianDataOrg/open-christian-data](https://huggingface.co/datasets/OpenChristianDataOrg/open-christian-data) is live. The release contains **408 work units**, **1,812 source files**, **544,961 top-level source records**, **1,811 exported source files**, and **805,151 flattened JSONL rows**. The verified data-upload revision immediately after all 12 JSONL uploads was `c2b4c0e0dd6676020ee089139b665383a5bc498e` at `2026-07-17T02:41:41+00:00`. The follow-up card-only correction does not alter this verified data payload; no follow-up card-only upload has happened yet.
80
+
81
+ All content is written by authors who died before 1928. All texts are in the public domain. The dataset is CC0.
82
+
83
+ ## Export shape
84
+
85
+ The live whole-corpus HuggingFace export is the per-schema JSONL export generated by
86
+ `build/scripts/export_huggingface.py`. Each published config corresponds to one schema-specific
87
+ JSONL file.
88
+
89
+ ## What's included
90
+
91
+ The public inventory is counted by recognized work units: the title-level item a reader, library, citation, or source catalog would treat as present. Export records, verses, dictionary entries, topics, Q&A pairs, chapters, and quotations are subordinate technical units.
92
+
93
+ Generated from current `data/` with `py -3 build/tools/count_dataset_records.py`.
94
+
95
+ | Category | Work units | Top-level records |
96
+ |---|---:|---:|
97
+ | Bible text | 9 | 282,395 |
98
+ | Catechisms | 15 | 3,509 |
99
+ | Church Fathers quotation source | 1 | 70,164 |
100
+ | Commentaries | 34 | 115,069 |
101
+ | Devotionals | 2 | 1,464 |
102
+ | Doctrinal documents | 38 | 38 |
103
+ | Hymns | 1 | 34,904 |
104
+ | Lexicon | 1 | 0 |
105
+ | Prayers | 4 | 205 |
106
+ | Reference works | 7 | 26,305 |
107
+ | Sermons | 7 | 5,297 |
108
+ | Structured texts | 288 | 289 |
109
+ | Topical reference works | 1 | 5,322 |
110
+ | **Total** | **408** | **544,961** |
111
+
112
+ The live HuggingFace export flattens structured texts and doctrinal documents, so this release contains **805,151 JSONL rows**. The source repository includes the generated work metadata audit at `docs/WORK_CATALOG.md` and a browser-friendly review surface at `docs/WORK_CATALOG.html`.
113
+
114
+ ## Loading the data
115
+
116
+ ```python
117
+ from datasets import load_dataset
118
+
119
+ # Load published configs
120
+ commentary = load_dataset("OpenChristianDataOrg/open-christian-data", "commentary")
121
+ structured_text = load_dataset("OpenChristianDataOrg/open-christian-data", "structured_text")
122
+ ```
123
+
124
+ ## Data structure
125
+
126
+ Every record is a flattened row combining the resource-level metadata with entry-level fields. Key fields present on all records:
127
+
128
+ | Field | Description |
129
+ |---|---|
130
+ | `_source_id` | Unique identifier for the parent resource (e.g., `matthew-henry-complete`) |
131
+ | `entry_id` | Unique identifier for this entry (e.g., `matthew-henry-complete.Ezek.1.1-3`) |
132
+ | `author` | Author name |
133
+ | `tradition` | Array of theological tradition labels (e.g., `["reformed", "puritan"]`) |
134
+ | `era` | Historical era (`patristic`, `medieval`, `reformation`, `post-reformation`, `modern`) |
135
+ | `license` | `cc0-1.0` or `public-domain` |
136
+ | `schema_type` | Content type — one of the config names above |
137
+
138
+ Schema-specific fields vary by content type. Full field documentation is in the [source repository](https://github.com/OpenChristianData/open-christian-data/blob/main/docs/SCHEMA_SPEC.md).
139
+
140
+ ## Provenance
141
+
142
+ Every record carries full provenance — the source URL, source format, download date, processing script, and a SHA-256 hash of the source file. This is stored in the `provenance_*` fields in the flattened export.
143
+
144
+ ## Verse references
145
+
146
+ All verse references use [OSIS format](https://crosswire.org/osis/) (`Gen.1.1`, `Rom.9.1-Rom.9.5`). Where sources use non-standard reference formats, both the raw form and the normalized OSIS form are stored.
147
+
148
+ ## License
149
+
150
+ - **Data**: [CC0 1.0 Universal](https://creativecommons.org/publicdomain/zero/1.0/) — all underlying texts are public domain; the structuring and provenance tracking are dedicated to the public domain
151
+ - **Bible text (BSB)**: [CC0 1.0](https://berean.bible/licensing.htm)
152
+ - **Code and schemas**: [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) — free for non-commercial use; commercial use requires a license (see [source repository](https://github.com/OpenChristianData/open-christian-data))
153
+
154
+ ## Citation
155
+
156
+ ```bibtex
157
+ @dataset{open_christian_data_2026,
158
+ title = {Open Christian Data},
159
+ author = {OpenChristianData},
160
+ year = {2026},
161
+ publisher = {HuggingFace},
162
+ url = {https://huggingface.co/datasets/OpenChristianDataOrg/open-christian-data},
163
+ license = {CC0-1.0}
164
+ }
165
+ ```
166
+
167
+ ## Source repository
168
+
169
+ [github.com/OpenChristianData/open-christian-data](https://github.com/OpenChristianData/open-christian-data)