Modified the im2latex page to include a link to prebuilt dataset with formulas taken from arXiv ( more accurately: http://www.cs.cornell.edu/projects/kddcup/datasets.html ). Also added link to github repo with tools used to generate the dataset.
This way people could try going straight into developing models instead of spending time gathering proper dataset.
Modified the im2latex page to include a link to prebuilt dataset with formulas taken from arXiv ( more accurately: http://www.cs.cornell.edu/projects/kddcup/datasets.html ). Also added link to github repo with tools used to generate the dataset. This way people could try going straight into developing models instead of spending time gathering proper dataset.
Link to dataset: https://zenodo.org/record/56198#.V2p0KTXT6eA Link to repo: https://github.com/Miffyli/im2latex-dataset
PS: I could be more familiar with Github's systems, and I apologize if I did this pull request wrong way around or so.