Enquiry into Objectives calculation for saving best model

pfnet-research / meta-tasnet

A PyTorch implementation of Meta-TasNet from "Meta-learning Extractors for Music Source Separation

MIT License

136 stars 16 forks source link

Closed mckhang closed 3 years ago

mckhang commented 3 years ago

Hi, thanks for your great work. Just one point of enquiry in objectives calculation prior to saving the best validation model in train.py

I am specifically referring to the following: objective = average_stats[-5:-1].mean()

I would like to ask why the slicing is [-5:-1] instead of [-4:] ?

Hope to seek your clarification and thanks in advance!

davda54 commented 3 years ago

Hi, thank you for finding this bug! I believe you're right, but changing it shouldn't make any significant difference :)