search

allegro / allRank

allRank is a framework for training learning-to-rank neural models based on PyTorch.
Apache License 2.0
850 stars 118 forks source link

Shape of the input of the Ordinal Loss #20

Closed mohuzi closed 3 years ago

mohuzi commented 3 years ago

https://github.com/allegro/allRank/blob/c90b64a0c46e3f9101524f9c31533ecf1ebc8384/allrank/models/losses/ordinal.py#L28 Shouldn't the shape be [batch_size, slate_length, n] in here? Thanks,

PrzemekPobrotyn commented 3 years ago

Yes, I believe you're right! Thanks for spotting this, will fix it soon.

sadaharu-inugami commented 3 years ago

Solved in a32a9bc. If that's all, you can close the issue now.