ArXiv parser
#2
by
chawins
- opened
Hi, thank you for maintaining this dataset! I was looking for something like this exactly (nicely parsed and dated text data), and I'm very happy to find this dataset.
Out of curiosity, I'm wondering about how arxiv papers are parsed into the text format here. I've been using Nougat to some success but less than ideal so I'm pretty curious about what others are using.
I am dealing with latex source instead of the PDFs. And you can find the code for the monthly updating here: https://github.com/liyucheng09/LatestEval/blob/master/arxiv_downloader.py
liyucheng
changed discussion status to
closed
Ah I see. Very cool. Thank you.