Change structure to welcome multilingual later

README.md

@@ -13,57 +13,17 @@ configs:
 - config_name: default
   data_files:
   - split: train
-    path: 20231220/*
-- config_name: "20231220"
+    path: "*/20231220/*.parquet"
+- config_name: "fr"
   data_files:
   - split: train
-    path: 20231220/*
-- config_name: "20231220.1"
+    path: "fr/20231220/*.parquet"
+- config_name: "sample"
   data_files:
   - split: train
     path: 20231220/train-000000-of-000007.parquet
-- config_name: "20231220.2"
-  data_files:
-  - split: train
-    path: 20231220/train-000001-of-000007.parquet
-- config_name: "20231220.3"
-  data_files:
-  - split: train
-    path: 20231220/train-000002-of-000007.parquet
-- config_name: "20231220.4"
-  data_files:
-  - split: train
-    path: 20231220/train-000003-of-000007.parquet
-- config_name: "20231220.5"
-  data_files:
-  - split: train
-    path: 20231220/train-000004-of-000007.parquet
-- config_name: "20231220.6"
-  data_files:
-  - split: train
-    path: 20231220/train-000005-of-000007.parquet
-- config_name: "20231220.7"
-  data_files:
-  - split: train
-    path: 20231220/train-000006-of-000007.parquet
 dataset_info:
-- config_name: default
-  features:
-  - name: id
-    dtype: int32
-  - name: url
-    dtype: string
-  - name: title
-    dtype: string
-  - name: text
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 442573568
-    num_examples: 650087
-  download_size: 192941244
-  dataset_size: 442573568
-- config_name: "20231220"
+- config_name: "fr"
   features:
   - name: id
     dtype: int32
@@ -79,7 +39,7 @@ dataset_info:
     num_examples: 650087
   download_size: 192941244
   dataset_size: 442573568
-- config_name: "20231220.1"
+- config_name: "sample"
   features:
   - name: id
     dtype: int32
@@ -95,126 +55,34 @@ dataset_info:
     num_examples: 91996
   download_size: 32850743
   dataset_size: 68635508
-- config_name: "20231220.2"
-  features:
-  - name: id
-    dtype: int32
-  - name: url
-    dtype: string
-  - name: title
-    dtype: string
-  - name: text
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 70148053
-    num_examples: 118924
-  download_size: 29931642
-  dataset_size: 70148053
-- config_name: "20231220.3"
-  features:
-  - name: id
-    dtype: int32
-  - name: url
-    dtype: string
-  - name: title
-    dtype: string
-  - name: text
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 73099864
-    num_examples: 119995
-  download_size: 30711503
-  dataset_size: 73099864
-- config_name: "20231220.4"
-  features:
-  - name: id
-    dtype: int32
-  - name: url
-    dtype: string
-  - name: title
-    dtype: string
-  - name: text
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 80580684
-    num_examples: 110066
-  download_size: 36351584
-  dataset_size: 80580684
-- config_name: "20231220.5"
-  features:
-  - name: id
-    dtype: int32
-  - name: url
-    dtype: string
-  - name: title
-    dtype: string
-  - name: text
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 74534373
-    num_examples: 102588
-  download_size: 31438062
-  dataset_size: 74534373
-- config_name: "20231220.6"
-  features:
-  - name: id
-    dtype: int32
-  - name: url
-    dtype: string
-  - name: title
-    dtype: string
-  - name: text
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 69491254
-    num_examples: 98966
-  download_size: 28958890
-  dataset_size: 69491254
-- config_name: "20231220.7"
-  features:
-  - name: id
-    dtype: int32
-  - name: url
-    dtype: string
-  - name: title
-    dtype: string
-  - name: text
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 6083832
-    num_examples: 7552
-  download_size: 2698820
-  dataset_size: 6083832
 ---
 
 # Plain text of French Wiktionary
-                
-This dataset was created by [OpenLLM France](https://openllmfrance.org/) from the [Wikimedia dumps](https://dumps.wikimedia.org/other/enterprise_html/runs).
 
-It is a plain text version of pages from [fr.wiktionary.org/wiki](https://fr.wiktionary.org/wiki)
-without HTML tags nor wiki templates.              
-The text just includes markdown syntax for headers, lists and tables.
-See [Notes on data formatting](#notes-on-data-formatting) for more details.
-                
-* [Statistics](#statistics)
-* [Example use (python)](#example-use-python)
-* [Data fields](#data-fields)
-* [Notes on data formatting](#notes-on-data-formatting)
+* [Dataset Description](#dataset-description)
+  * [Size](#size)
+  * [Example use (python)](#example-use-python)
+  * [Data fields](#data-fields)
+  * [Notes on data formatting](#notes-on-data-formatting)
 * [License](#license)
 * [Aknowledgements](#aknowledgements)
 * [Citation](#citation)
 
+## Dataset Description
+
+This dataset is a plain text version of pages from [wiktionary.org](https://wiktionary.org/) in French language.
+The text is without HTML tags nor wiki templates.
+It just includes markdown syntax for headers, lists and tables.
+See [Notes on data formatting](#notes-on-data-formatting) for more details.
+
+It was created by [LINAGORA](https://labs.linagora.com/) and [OpenLLM France](https://openllmfrance.org/)
+from the [Wikimedia dumps](https://dumps.wikimedia.org/other/enterprise_html/runs),
+using code in [https://github.com/OpenLLM-France/wikiplaintext](https://github.com/OpenLLM-France/wikiplaintext).
                 
-## Statistics
+### Size
                 
 The amount of data for the latest dump (20231220) is:
-| **Unit**        | **Size** |
+|                 | **French** |
 | :---------------|--------: |
 | # documents     |   650 087 |
 | # paragraphs    |   3 410 374 |
@@ -224,33 +92,28 @@ The amount of data for the latest dump (20231220) is:
 | size in memory  |   422.1 MB |
 | size on disk    |   184.0 MB |
 
-## Example use (python)
+### Example use (python)
 
 Load the full dataset:
 
 ```python
 import datasets
 
-ds = datasets.load_dataset("OpenLLM-France/wiktionary.fr")
+ds = datasets.load_dataset("OpenLLM-France/wiktionary.fr", streaming=True, split="train")
 ```
 
 Load only a small subset:
 ```python
-import datasets
-
-i = 1 # Can be a number between 1 and 7
-subset = f"20231220.{i}"
-
-ds = datasets.load_dataset("OpenLLM-France/wiktionary.fr", subset)
+ds = datasets.load_dataset("OpenLLM-France/wiktionary.fr", "sample", split="train")
 ```
 
 ## Data fields
 
 The data fields are the same among all configurations:
-- `id` (int): ID of the article.
-- `url` (`str`): URL of the article.
-- `title` (`str`): Title of the article.
-- `text` (`str`): Text content of the article.
+- `id` (int): ID of the page.
+- `url` (`str`): URL of the page.
+- `title` (`str`): Title of the page.
+- `text` (`str`): Text content of the page.
 
 For example:
 ```
@@ -263,7 +126,7 @@ For example:
 
 ## Notes on data formatting
 
-see [OpenLLM-France/wikipedia.fr](https://huggingface.co/datasets/OpenLLM-France/wikipedia.fr)
+see [OpenLLM-France/wikipedia.fr](https://huggingface.co/datasets/OpenLLM-France/wikipedia)
 
 ## License
 
@@ -271,7 +134,9 @@ This dataset is distributed under the [Creative Commons Attribution-ShareAlike 4
 
 ## Aknowledgements
 
-This dataset was created by [Jérôme Louradour](https://huggingface.co/Jeronymous) on behalf of [OpenLLM France](https://openllmfrance.org/).
+This dataset was created by [Jérôme Louradour](https://huggingface.co/Jeronymous) on behalf of
+[LINAGORA](https://labs.linagora.com/) and
+[OpenLLM France](https://openllmfrance.org/).
 
 Many thanks to the [Wikimedia Foundation](https://wikimediafoundation.org/)
 for providing the data and useful advices,
@@ -280,10 +145,10 @@ in particular Isaac Johnson, Albert Villanova and Rémy Gerbet.
 ## Citation
 
 ```
-@online{wikipedia_fr_dump,
-    author = "OpenLLM France",
-    title  = "Plain text of French Wiktionary",
-    url    = "https://huggingface.co/datasets/OpenLLM-France/wiktionary.fr"
+@online{wiktionary_fr_dump,
+    author = "Jérôme Louradour, OpenLLM-France, LINAGORA Labs",
+    title  = "Plain text of Wiktionary",
+    url    = "https://huggingface.co/datasets/OpenLLM-France/wiktionary"
 }
 ```