Update README.md

README.md

@@ -309,12 +309,19 @@ configs:
   data_files:
   - split: train
     path: wikimedia_others/train-*
+task_categories:
+- text-classification
+- text-generation
+language:
+- it
+size_categories:
+- 100B<n<1T
 ---
 WARNING: THIS "README" IS JUST A STUB, IT WILL BE IMPROVED DURING THE
 NEXT FEW DAYS AND FILLED WITH MANY OTHER
 INFORMATIONS AND DETAILED STATISTICS
 
-**Testimole -- A multi-billion Italian text corpus **
+**Testimole -- A multi-billion tokens Italian text corpus**
 
 The goal of this work is to create a huge linguistic resource for the
 Italian language that can be used for several NLP applications,
@@ -350,8 +357,7 @@ goals.
 This is the project that started the entire work: the goal was to
 collect the largest possible amount of Usenet posts published in the
 hierachies it.\* and italia.\* \[3\], as they were listed on
-"[www.eternal-september.org](http://www.eternal-september.org/)" \[4\]
-and gathered mainly from Google Groups archive.
+"www.eternal-september.org" and gathered mainly from Google Groups archive.
 
 This split contains 19.395.579.455 tokens. Texts were not checked for
 language, but it is a safe assumption that most of the text contained is
@@ -360,21 +366,18 @@ in Italian as the selected Usenet hierarchies target only Italian users.
 Detailed statistics, already computed, will follow very soon. For now,
 here are general stats about this part of the dataset:
 
-{
 
-\"char\_count\": 59389804791,
 
-\"token\_count\": 19395579455,
+\"chars": 59389804791,
 
-\"sent\_count\": 519535427,
+\"tokens": 19395579455,
 
-\"post\_count\": 89499446,
+\"sentences": 519535427,
 
-\"thread\_count\": 14521548,
+\"post": 89499446,
 
-\"author\_count\": 3626546,
+\"thread": 14521548,
 
-}
 
 83GB of JSONL file before the conversion to HuggingFace dataset
 
@@ -394,19 +397,15 @@ general stats about this part of the dataset:
 
 {
 
-\"char\_count\": 199436329709,
-
-\"token\_count\": 62415825978,
+\"chars": 199436329709,
 
-\"text\_bytes\": 201359756617,
+\"tokens": 62415825978,
 
-\"sent\_count\": 1673025712,
+\"sentences": 1673025712,
 
-\"post\_count\": 468391746,
+\"posts": 468391746,
 
-\"thread\_count\": 25280745,
-
-\"author\_count\": 37426524,
+\"threads": 25280745,
 
 \"hasImage\": 46071
 
@@ -621,11 +620,11 @@ contain html code.
 *Work in progress; this will contain statistics of tokens, chars and sentences lenght for each diachronic resource (usenet newsgroup, post, blog) for each month for each year*
 
 **Conclusions**
-
+*Work in progress*
 **References (partial)**
 
 \* \[1\] <https://pdai.info/>
 
 \* \[2\] https://github.com/openai/tiktoken
 
-\* \[3\] <https://xmau.com/usenet/>
+\* \[3\] <https://xmau.com/usenet/>