Update README.md

README.md

@@ -4,19 +4,29 @@ language:
 license:
 - unknown
 size_categories:
-- 10K<n<100K
+- 1M<n<10M
 task_categories:
 - token-classification
 tags:
 - ner
+- DFP
+- french prompts
+annotations_creators:
+- found
+language_creators:
+- found
+multilinguality:
+- monolingual
+source_datasets:
+- wikineural
 ---
 
 # wikineural_fr_prompt_ner
 ## Summary
 
-**wikineural_fr_prompt_ner** is a subset of the [**Dataset of French Prompts (DFP)**]().
-It contains **X** rows that can be used for a name entity recognition task.
-The original data (without prompts) comes from the dataset [wikineural](https://huggingface.co/datasets/tner/wikineural) by Tedeschi et al. where only the French part has been kept.
+**wikineural_fr_prompt_ner** is a subset of the [**Dataset of French Prompts (DFP)**](https://huggingface.co/datasets/CATIE-AQ/DFP).  
+It contains **2,647,638** rows that can be used for a name entity recognition task.  
+The original data (without prompts) comes from the dataset [wikineural](https://huggingface.co/datasets/tner/wikineural) by Tedeschi et al. where only the French part has been kept.  
 A list of prompts (see below) was then applied in order to build the input and target columns and thus obtain the same format as the [xP3](https://huggingface.co/datasets/bigscience/xP3) dataset by Muennighoff et al.
 
 
@@ -59,9 +69,9 @@ wikineural['train']['tags'] = list(map(lambda x: x.replace("[","").replace("]","
 
 
 # Splits
-- train with X samples
-- dev with Y samples
-- test with Z samples
+- `train` with 2,116,800 samples
+- `valid` with 264,600 samples
+- `test` with 266,238 samples
 
 
 # How to use?