[Private 2nd, 0.9837], Part 1 of 2, Summary, Data creation
2nd place solution, Part 1 of 2
May I ask are there any additional dataset that you have utilized? Our team trained more than 900K for each folds, but failed to attain such performance with the same PLM 😂
I did not use any additional data. One possible reason is different data cleaning.
Did converting Python2 to Python3 code boost the classification performance? Haven't tested that one yet!
I haven't tested it specifically, but I think it doesn't make much of a difference.
Thank you for the answer 🤗
You're welcome. Congrats on a high spot in the leaderboard.