Update README.md
Browse files
README.md
CHANGED
|
@@ -40,6 +40,8 @@ configs:
|
|
| 40 |
* **GitHub:** [https://github.com/marcodsn/academic-chains](https://github.com/marcodsn/academic-chains/tree/main/large_scale)
|
| 41 |
* **Dataset:** [https://huggingface.co/datasets/marcodsn/arxiv-markdown](https://huggingface.co/datasets/marcodsn/arxiv-markdown) (this page)
|
| 42 |
|
|
|
|
|
|
|
| 43 |
This dataset contains open-access papers retrieved from [arXiv](https://arxiv.org) and converted to markdown format using [docling](https://github.com/docling-project/docling); specifically, we use the following docling pipeline:
|
| 44 |
```python
|
| 45 |
pipeline_options = PdfPipelineOptions()
|
|
|
|
| 40 |
* **GitHub:** [https://github.com/marcodsn/academic-chains](https://github.com/marcodsn/academic-chains/tree/main/large_scale)
|
| 41 |
* **Dataset:** [https://huggingface.co/datasets/marcodsn/arxiv-markdown](https://huggingface.co/datasets/marcodsn/arxiv-markdown) (this page)
|
| 42 |
|
| 43 |
+
> **[24/04/2025]** Because of the high storage utilization we are seeing, we are currently thinking about moving the embedded images to a [Cloudflare R2](https://www.cloudflare.com/developer-platform/products/r2/) instance and to only reference the URLs in the markdowns, stay tuned for more updates on this!
|
| 44 |
+
|
| 45 |
This dataset contains open-access papers retrieved from [arXiv](https://arxiv.org) and converted to markdown format using [docling](https://github.com/docling-project/docling); specifically, we use the following docling pipeline:
|
| 46 |
```python
|
| 47 |
pipeline_options = PdfPipelineOptions()
|