add readme

README.md

@@ -27,3 +27,41 @@ configs:
   - split: train
     path: data/train-*
 ---
+## Arxiver Dataset
+Arxiver consists of 138,830 [arXiv](https://arxiv.org/) papers converted to multi-markdown (**.mmd**) format. Our dataset includes original arXiv article IDs, titles, abstracts, authors, publication dates, URLs and corresponding markdown files published between January 2023 and October 2023.
+
+We hope our dataset will be useful for various applications such as semantic search, domain specific language modeling, question answering and summarization.
+
+## Curation
+The Arxiver dataset is created using a neural OCR - [Nougat](https://facebookresearch.github.io/nougat/). After OCR processing, we apply custom text processing steps to refine the data. This includes extracting author information, removing reference sections, and performing additional cleaning and formatting.
+
+## Using Arxiver
+You can easily download and use the arxiver dataset with Hugging Face's [datasets](https://huggingface.co/datasets) library.
+```py
+from datasets import load_dataset
+
+# whole dataset takes 3.3GB
+dataset = load_dataset("neuralwork/arxiver") 
+print(dataset)
+```
+
+Alternatively, you can stream the dataset to save disk space or to partially download the dataset:
+```py
+from datasets import load_dataset
+
+dataset = load_dataset("neuralwork/arxiver", streaming=True)
+print(dataset)
+print(next(iter(dataset['train'])))
+```
+
+## References
+The original articles are maintained by [arXiv](https://arxiv.org/). If you use this dataset in your research or project, please cite it as follows:
+```
+@misc{acar_arxiver2024,
+  author = {Alican Acar, Alara Dirik, Muhammet Hatipoglu},
+  title = {ArXiver},
+  year = {2024},
+  publisher = {Hugging Face},
+  howpublished = {\url{https://huggingface.co/datasets/neuralwork/arxiver}}
+}
+```