Datasets:

Modalities:
Text
Formats:
json
ArXiv:
Libraries:
Datasets
pandas
License:

Duplicate Issue in the 72k Dataset

#4
by Smile2Play - opened

During the RL training process, I noticed that solve_equal was behaving abnormally. So, I conducted a duplication check and found that a large portion of the data was duplicated: 48,419 out of 72,444?

Open-Reasoner-Zero org

We conduct the issue for each dataset, and when combining the globally deduplication is actually lacked, and we were aware of this very recently (after the training is finished).
We are plan to release a better version of dataset recently, with improved data quality and will fix this issue. Please stay tuned!

reign12 changed discussion status to closed
Your need to confirm your account before you can post a new comment.

Sign up or log in to comment