Update README.md

README.md

@@ -20,8 +20,6 @@ configs:
   data_files:
   - split: train
     path: data/train-*
----
----
 pretty_name: github-java-corpus
 language:
 - code
@@ -34,7 +32,7 @@ tags:
 - java
 ---
 
-# github-java-2012
+# github-java-corpus
 
 ## Summary
 This dataset contains Java source-code text samples prepared for pretraining.
@@ -60,4 +58,24 @@ The dataset was built from streamed archive processing into file-level text samp
 ## Notes
 - Intended for pretraining
 - One row corresponds to one source file
-- `Date` is the assigned year label for this source
+- `Date` is the assigned year label for this source
+
+## Source Dataset
+
+This dataset is derived from the Project CodeNet dataset.
+
+Original dataset:
+https://datashare.ed.ac.uk/handle/10283/2334
+
+Project CodeNet is a large-scale dataset of source code collected from online programming platforms.
+
+## Citation
+
+If you use this dataset, please cite the original Project CodeNet dataset:
+
+@inproceedings{puri2021project,
+  title={Project CodeNet: A Large-Scale AI for Code Dataset for Learning a Diversity of Coding Tasks},
+  author={Puri, Ruchir and Kung, David and Janssen, Geert and others},
+  booktitle={NeurIPS Datasets and Benchmarks},
+  year={2021}
+}