Update README.md

README.md

@@ -1,3 +1,96 @@
----
-license: mit
----
+---
+license: mit
+size_categories:
+- 1M<n<10M
+task_categories:
+- text-generation
+language:
+- en
+---
+# Dataset Card for Dataset Name
+
+***OpenBrt-2m***
+
+## Dataset Details
+
+### Dataset Description
+
+* This dataset;
+* Contains 3 files:
+
+* train.csv = Not filtered file, contains: 2.4m tokens, avarage qscore (quality score) = ~0.5
+* train_filtered.csv = Filtered with **quality score** (Only contains: qscore > 0.7 random articles from Wikipedia), avarage qscore (quality score) = ~1.058
+* train_ultra_filtered.csv = Filtered with high **quality score** (Only contains: qscore > 1.0 random articles from Wikipedia.), avarage qscore (quality score) = ~1.16
+
+
+- **Curated by:** Bertug Gunel (Bertuğ Günel)
+- **Funded by [optional]:** Nobody
+- **Shared by [optional]:** Nobody
+- **Language(s) (NLP):** EN (en, En, English)
+- **License:** MIT (Mit, mit)
+
+### Dataset Sources [optional]
+
+<!-- Provide the basic links for the dataset. -->
+
+- **Repository:** Cooming soon!
+- **Paper [optional]:** Cooming soon!
+- **Demo [optional]:** Cooming soon!
+
+## Uses
+
+<!-- Address questions around how the dataset is intended to be used. -->
+
+### Direct Use
+
+***You can install ".csv" files and use it!
+
+### Out-of-Scope Use
+
+Model is not good for "Daily Dialogues" 
+
+## Dataset Structure
+
+Only train split, 5mb (biggest file in the split)
+## Dataset Creation
+
+### Curation Rationale
+
+* We trained a model but datasets quality score is about: 0.4 and not good at STEM (Math, science, history, code, chemist, ceography etc.)
+* Then we want to create our dataset with Wikipedia! 
+
+### Source Data
+
+Randomly selected Wikipedia articles. (230x, about 2.4 million token in English)
+
+#### Data Collection and Processing
+- We collected,
+- We filtered "x" to 'x'
+- We filtered bad punctuation (except punctuation marks like {[]}\|-!'^+%&/()=?*_.,;:<>"")
+- We filtered low quality tokens/sentences/articles
+
+#### Who are the source data producers?
+
+- Source is: Wikipedia (c.c. licence)
+- Producers: All the writers and users of Wikipedia
+
+
+#### Personal and Sensitive Information
+
+- Dataset is NOT CONTAIN any SPECIAL or SENSETIVE data.
+
+- AT YOUR OWN RISK
+
+## Bias, Risks, and Limitations
+
+- Dataset may contain data ***NOT SUITABLE FOR ALL AUDIENCES***, such as ***political***, ***sexual***, ***18+***, ***gambling***, ***betting***, ***drugs***, ***violence***, ***horror***, ***blood***, **descriptions of ***illegal*** activities!**
+
+- ***USE AT YOUR OWN RISK***
+
+
+## Dataset Card Contact
+
+* Bertug Gunel,
+* Turkey,
+* Eskisehir,
+* bertugscpmail@gmail.com or bertug2099@gmail.com for contact, you can contact FREELY!