Harm-P and Harm-C Datasets
adrielkuek opened this issue · 2 comments
adrielkuek commented
Hi, I'm trying to reproduce the results for the Momenta framework as seen from the paper. Understand that the HarMeme dataset is a large-scale dataset of close to almost 10k memes. However the links for the gdrive download to the 2 datasets only has 150 meme images each. Would it be possible to release to entire HarMeme dataset for testing?
Thanks and appreciate it!
ngthanhtin commented
Maybe after their paper got accepted, they will publish the full dataset.
But have you fixed the wrong indexes problem in JSON files? As I stated here: #2
shiv6891 commented
The complete dataset links are now added to the main readme.