Update README.md

README.md

@@ -16,7 +16,7 @@ dataset_info:
     num_bytes: 3717538.0422429685
     num_examples: 2375
   download_size: 224184711
-  dataset_size: 371875896.0
+  dataset_size: 371875896
 - config_name: 100k
   features:
   - name: text
@@ -33,7 +33,7 @@ dataset_info:
     num_bytes: 3654.5346534653463
     num_examples: 3
   download_size: 212072
-  dataset_size: 369108.0
+  dataset_size: 369108
 - config_name: 10B
   features:
   - name: text
@@ -50,7 +50,7 @@ dataset_info:
     num_bytes: 103964370.23416495
     num_examples: 64000
   download_size: 25249998174
-  dataset_size: 40008285333.0
+  dataset_size: 40008285333
 - config_name: 10M
   features:
   - name: text
@@ -67,7 +67,7 @@ dataset_info:
     num_bytes: 373730.00803400803
     num_examples: 256
   download_size: 22486785
-  dataset_size: 37432856.0
+  dataset_size: 37432856
 - config_name: 10k
   features:
   - name: text
@@ -101,7 +101,7 @@ dataset_info:
     num_bytes: 104972711.86121707
     num_examples: 64000
   download_size: 37966833792
-  dataset_size: 60119681222.0
+  dataset_size: 60119681222
 - config_name: 1B
   features:
   - name: text
@@ -118,7 +118,7 @@ dataset_info:
     num_bytes: 38437701.880162396
     num_examples: 28692
   download_size: 2346974411
-  dataset_size: 3843814397.0
+  dataset_size: 3843814397
 - config_name: 1M
   features:
   - name: text
@@ -135,7 +135,7 @@ dataset_info:
     num_bytes: 37019.21197648787
     num_examples: 27
   download_size: 2183019
-  dataset_size: 3732085.0
+  dataset_size: 3732085
 - config_name: 20B
   features:
   - name: text
@@ -152,7 +152,7 @@ dataset_info:
     num_bytes: 105482877.0574707
     num_examples: 64000
   download_size: 50682523292
-  dataset_size: 80231072356.0
+  dataset_size: 80231072356
 - config_name: 25B
   features:
   - name: text
@@ -169,7 +169,7 @@ dataset_info:
     num_bytes: 105790923.98284689
     num_examples: 64000
   download_size: 63397565382
-  dataset_size: 100342468245.0
+  dataset_size: 100342468245
 - config_name: 30B
   features:
   - name: text
@@ -186,7 +186,7 @@ dataset_info:
     num_bytes: 105997103.53253783
     num_examples: 64000
   download_size: 76111936677
-  dataset_size: 120453859676.0
+  dataset_size: 120453859676
 - config_name: 5B
   features:
   - name: text
@@ -203,7 +203,7 @@ dataset_info:
     num_bytes: 101031980.90819068
     num_examples: 64000
   download_size: 12526141470
-  dataset_size: 19896889444.0
+  dataset_size: 19896889444
 configs:
 - config_name: 100M
   data_files:
@@ -277,11 +277,22 @@ configs:
     path: 5B/train-*
   - split: test
     path: 5B/test-*
+task_categories:
+- text-generation
+- text2text-generation
 ---
 
+# Filtered CulturaX + Wikipedia for Dutch
+
+This is a combined and filtered version of [CulturaX](https://huggingface.co/datasets/uonlp/CulturaX) and [Wikipedia](https://huggingface.co/datasets/wikimedia/wikipedia), only including Dutch. It is intended for the training of LLMs.
+
+Different configs are available based on the number of tokens (see a section below with an overview). This can be useful if you want to know exactly how many tokens you have. Great for using as a streaming dataset, too. Tokenization is done with the big vocabulary of the `google/gemma-2b` tokenizer so depending on your tokenizer these exact numbers may differ.
+
 
 ## Filtering
 
+While CultruaX already has done a lot of filtering, some more filtering can be done to improve the quality of the corpus. These filters are described below.
+
 The baseline ratios (punctuation, uppercase, digits) were calculated on the SONAR-500 corpus (excluding WRPEA WRPED WRUEA WRUED WRUEB).
 
 **CulturaX**:
@@ -531,4 +542,4 @@ BAD_PHRASES_DOC_LEVEL = {
 - test_num_tokens: 26,520,069
 - total_num_samples: 84,753,828
 - train_num_samples: 84,689,828
-- test_num_samples: 64,000
+- test_num_samples: 64,000