Update README.md

README.md

@@ -8,35 +8,32 @@ language:
 ---
 # C4 English Tokenized Samples
 
-This dataset contains tokenized English samples from the C4 (Colossal Clean Crawled Corpus) dataset. It provides a preprocessed subset of the original C4 data, making it easier to use for various natural language processing tasks.
+This dataset contains tokenized English samples from the C4 (Colossal Clean Crawled Corpus) dataset for natural language processing tasks.
 
-## Dataset Details
-
-- **Source**: First 110 000 entries from the `en` split of [allenai/c4](https://huggingface.co/datasets/allenai/c4)
-- **Preprocessing**:
-  1. Tokenized using [spaCy](https://spacy.io/)'s `en_core_web_sm` model
-  2. Lowercased
-  3. Tokens joined with spaces
+The first 125 000 entries from the `en` split of [allenai/c4](https://huggingface.co/datasets/allenai/c4)
+were tokenized using [spaCy](https://spacy.io/)'s `en_core_web_sm` model. Tokens joined with spaces.
 
 ## Features
 
 - `text`: Original text from C4
-- `tokenized`: Preprocessed text (tokenized, lowercased, and space-joined)
+- `tokenized`: The tokenized and space-joined text
 - `num_tokens`: Number of tokens after tokenization
+- `num_punct_tokens`: Number of punctuation tokens after tokenization
 
 ## Example
 
 ```json
 {
-  "text": "The Denver Board of Education opened the 2017-18 school year with an update on projects that include new construction, upgrades, heat mitigation [...]",
-  "tokenized": "the denver board of education opened the 2017 - 18 school year with an update on projects that include new construction , upgrades , heat mitigation [...]",
-  "num_tokens": 192
+  "text": "ALDUS MANUTIUS AND HIS THESAURUS CORNUCOPIAE OF 1496.\nSyracuse (1958) . 7.5 x 4.25, cloth, 32 pp, a v.g. copy [...]",
+  "tokenized": "ALDUS MANUTIUS AND HIS THESAURUS CORNUCOPIAE OF 1496 . \n Syracuse ( 1958 ) . 7.5 x 4.25 , cloth , 32 pp , a v.g . copy [...]",
+  "num_tokens": 84,
+  "num_punct_tokens": 19
 }
 ```
 
 ## Usage
 
-This preprocessed dataset can be particularly useful for:
+This dataset can be useful for:
 - Text classification tasks
 - Language modeling
 - Sentiment analysis