Datasets:
Tasks:
Text Generation
Formats:
parquet
Sub-tasks:
language-modeling
Languages:
Danish
Size:
1M - 10M
License:
Kenneth Enevoldsen
commited on
ensure domains in metadata
Browse files
data/wikibooks/wikibooks.md
CHANGED
@@ -13,6 +13,8 @@ task_ids:
|
|
13 |
- language-modeling
|
14 |
source_datasets:
|
15 |
- danish-foundation-models/danish-gigaword
|
|
|
|
|
16 |
---
|
17 |
|
18 |
# Dataset Card for Wikibooks
|
|
|
13 |
- language-modeling
|
14 |
source_datasets:
|
15 |
- danish-foundation-models/danish-gigaword
|
16 |
+
domains:
|
17 |
+
- Encyclopedic
|
18 |
---
|
19 |
|
20 |
# Dataset Card for Wikibooks
|
data/wikisource/descriptive_stats.json
CHANGED
@@ -2,6 +2,5 @@
|
|
2 |
"number_of_samples": 2419,
|
3 |
"average_document_length": 6398.767259198015,
|
4 |
"number_of_tokens": 5344862,
|
5 |
-
"
|
6 |
-
"revision": "341a0162905ca4848fc840d941edfd0a6d1b2314"
|
7 |
}
|
|
|
2 |
"number_of_samples": 2419,
|
3 |
"average_document_length": 6398.767259198015,
|
4 |
"number_of_tokens": 5344862,
|
5 |
+
"revision": "f1205a6d2f118eeb1b5181e7281b3909963b0678"
|
|
|
6 |
}
|
data/wikisource/images/dist_document_length.png
CHANGED
![]() |
Git LFS Details
|
![]() |
Git LFS Details
|
data/wikisource/wikisource.md
CHANGED
@@ -1,19 +1,22 @@
|
|
1 |
---
|
2 |
pretty_name: Wikisource
|
3 |
language:
|
4 |
-
|
5 |
license: cc0-1.0
|
6 |
license_name: CC-0
|
7 |
size_categories:
|
8 |
-
|
9 |
task_categories:
|
10 |
-
|
11 |
-
|
12 |
task_ids:
|
13 |
-
|
14 |
source_datasets:
|
15 |
-
|
|
|
|
|
16 |
---
|
|
|
17 |
# Dataset Card for Wikisource
|
18 |
|
19 |
<!-- START-SHORT DESCRIPTION -->
|
@@ -26,6 +29,7 @@ The Danish subsection of [Wikisource](https://en.wikisource.org/wiki/Main_Page).
|
|
26 |
|
27 |
<!-- START-DESC-STATS -->
|
28 |
- **Language**: dan, dansk, Danish
|
|
|
29 |
- **Number of samples**: 2.42K
|
30 |
- **Number of tokens (Llama 3)**: 5.34M
|
31 |
- **Average document length (characters)**: 6398.77
|
@@ -54,22 +58,20 @@ An example from the dataset looks as follows.
|
|
54 |
|
55 |
An entry in the dataset consists of the following fields:
|
56 |
|
|
|
57 |
- `text`(`str`): The content of the document.
|
58 |
- `source` (`str`): The source of the document (see [Source Data](#source-data)).
|
59 |
-
- `id` (`str`): An unique identifier for each document.
|
60 |
- `added` (`str`): An date for when the document was added to this collection.
|
61 |
- `created` (`str`): An date range for when the document was originally created.
|
62 |
-
- `
|
63 |
-
- `domain` (`str`): The domain of the source
|
64 |
-
- `metadata/source-pretty` (`str`): The long form version of the short-form source name
|
65 |
-
- `metadata/*`: Potentially additional metadata
|
66 |
<!-- END-SAMPLE -->
|
67 |
|
68 |
### Dataset Statistics
|
69 |
|
70 |
<!-- START-DATASET PLOTS -->
|
|
|
71 |
<img src="./images/dist_document_length.png" width="600" style="margin-right: 10px;" />
|
72 |
-
|
73 |
<!-- END-DATASET PLOTS -->
|
74 |
|
75 |
|
|
|
1 |
---
|
2 |
pretty_name: Wikisource
|
3 |
language:
|
4 |
+
- da
|
5 |
license: cc0-1.0
|
6 |
license_name: CC-0
|
7 |
size_categories:
|
8 |
+
- 1-10k
|
9 |
task_categories:
|
10 |
+
- text-generation
|
11 |
+
- fill-mask
|
12 |
task_ids:
|
13 |
+
- language-modeling
|
14 |
source_datasets:
|
15 |
+
- danish-foundation-models/danish-gigaword
|
16 |
+
domains:
|
17 |
+
- Encyclopedic
|
18 |
---
|
19 |
+
|
20 |
# Dataset Card for Wikisource
|
21 |
|
22 |
<!-- START-SHORT DESCRIPTION -->
|
|
|
29 |
|
30 |
<!-- START-DESC-STATS -->
|
31 |
- **Language**: dan, dansk, Danish
|
32 |
+
- **Domains**: None
|
33 |
- **Number of samples**: 2.42K
|
34 |
- **Number of tokens (Llama 3)**: 5.34M
|
35 |
- **Average document length (characters)**: 6398.77
|
|
|
58 |
|
59 |
An entry in the dataset consists of the following fields:
|
60 |
|
61 |
+
- `id` (`str`): An unique identifier for each document.
|
62 |
- `text`(`str`): The content of the document.
|
63 |
- `source` (`str`): The source of the document (see [Source Data](#source-data)).
|
|
|
64 |
- `added` (`str`): An date for when the document was added to this collection.
|
65 |
- `created` (`str`): An date range for when the document was originally created.
|
66 |
+
- `token_count` (`int`): The number of tokens in the sample computed using the Llama 8B tokenizer
|
|
|
|
|
|
|
67 |
<!-- END-SAMPLE -->
|
68 |
|
69 |
### Dataset Statistics
|
70 |
|
71 |
<!-- START-DATASET PLOTS -->
|
72 |
+
<p align="center">
|
73 |
<img src="./images/dist_document_length.png" width="600" style="margin-right: 10px;" />
|
74 |
+
</p>
|
75 |
<!-- END-DATASET PLOTS -->
|
76 |
|
77 |
|
src/tests/test_datasheets.py
CHANGED
@@ -45,3 +45,11 @@ def test_datasheet_required_headings(dataset_name: str):
|
|
45 |
req_h2_headings = ["## Dataset Description", "## Additional Information"]
|
46 |
for req_h2 in req_h2_headings:
|
47 |
assert ds_sheet.get_section_by_header(req_h2)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
45 |
req_h2_headings = ["## Dataset Description", "## Additional Information"]
|
46 |
for req_h2 in req_h2_headings:
|
47 |
assert ds_sheet.get_section_by_header(req_h2)
|
48 |
+
|
49 |
+
|
50 |
+
@pytest.mark.parametrize("dataset_name", DATASET_NAMES)
|
51 |
+
def test_domains_in_frontmatter(dataset_name: str):
|
52 |
+
readme = repo_path / "data" / dataset_name / f"{dataset_name}.md"
|
53 |
+
ds_sheet = DataSheet.load_from_path(readme)
|
54 |
+
|
55 |
+
assert ds_sheet.domains, "domains annotations are missing"
|